B) More than 50%

Explanation: It is stated that more than half of US startups were unable to survive beyond three years during the specified period, highlighting the challenges faced by new businesses.

C) They accelerate faster than revenues

Explanation: As a venture grows, selling, general, and administrative (SG&A) expenses tend to accelerate faster than revenues, which can strain financial resources.

B) Response rate to the average response rate

Explanation: Lift is calculated as the ratio of the response rate for each decile to the average response rate, allowing for a clear comparison of performance across segments.

B) TP + TN / (TP + TN + FP + FN)

Explanation: Accuracy is calculated as the sum of True Positives and True Negatives divided by the total number of predictions (TP + TN + FP + FN), indicating the proportion of correct predictions.

C) $400

Explanation: Pets.com spent an acquisition cost of $400 per customer, which highlights the significant investment made in acquiring new customers through various advertising channels.

B) Logistic Regression

Explanation: Logistic Regression is one of the techniques covered in the class, which is commonly used for binary classification problems in marketing and other fields.

B) Data Preprocessing

Explanation: Data Preprocessing is a crucial initial step in developing an intelligent system, as it involves preparing and cleaning the data to ensure it is suitable for analysis and modeling.

C) To reduce the dimensionality of data

Explanation: Feature extraction aims to simplify the data by reducing its dimensionality while retaining essential information, making it easier to analyze and model.

B) t = (β2 - 0) / SE

Explanation: The T-statistic for a coefficient in simple linear regression is calculated as t = (β2 - 0) / SE, where β2 is the coefficient estimate and SE is the standard error.

B) It indicates the number of correct negative predictions

Explanation: True Negative (TN) represents the instances where the model correctly predicts the negative class, which is crucial for evaluating the model's performance in binary classification.

C) American Express

Explanation: American Express is mentioned as a company that increased its market size by allowing customers to withdraw cash while traveling, targeting a specific need of travelers.

B) To analyze, explore, and predict outcomes

Explanation: The main goal of a data mining task is to analyze and explore data to make predictions about future outcomes, which is essential for informed decision-making.

C) SR3

Explanation: The announcement specifies that SR3 will be used for both the lecture from Week 3 and the tutorial from Week 4, indicating the location for these sessions.

B) Pricing

Explanation: The course includes a focus on pricing, which is a critical aspect of business strategy and economic value, particularly in the context of intelligent systems.

B) Support Vector Machine

Explanation: Support Vector Machine is a method discussed in the class, known for its effectiveness in classification tasks, particularly in marketing applications.

B) To evaluate the effectiveness of the predictive model

Explanation: Calculating lift for each decile helps in assessing how well the predictive model performs across different segments, providing insights into its effectiveness.

E) True Neutral (TN)

Explanation: True Neutral is not a recognized component in binary classification metrics. The correct components are True Positive (TP), False Positive (FP), True Negative (TN), and False Negative (FN).

B) Youden’s index

Explanation: Youden’s index is one of the metrics mentioned that can be used to determine the optimal threshold for classification, alongside lift value, focusing on the model's financial performance.

B) Positive word-of-mouth

Explanation: The text specifically mentions that restaurants, movies, and other entertainment services rely heavily on positive word-of-mouth as a key source of awareness, impacting consumer decisions.

C) $20

Explanation: The incremental profit margin for each pet food purchase was $20, which is important for calculating the break-even point for customer acquisition.

C) Logistic regression

Explanation: Logistic regression is specifically designed for predicting binary outcomes and ensures that predicted probabilities remain within the valid range of 0 to 1, making it a more suitable choice than linear regression.

B) Targeting current customers

Explanation: The course IS4242, titled 'Intelligent Systems & Techniques,' specifically focuses on targeting current customers, indicating its relevance to marketing and customer relationship management.

C) The strength of evidence against the null hypothesis

Explanation: The p-value is a measure of evidence against the null hypothesis; a smaller p-value indicates stronger evidence against it.

B) Stronger evidence against the null hypothesis

Explanation: A smaller p-value suggests stronger evidence against the null hypothesis, indicating that the observed data is less likely under the null hypothesis.

C) It generates awareness and impacts purchasing intention

Explanation: Positive word-of-mouth is highlighted as a significant factor that generates awareness and influences consumers' intentions to purchase, particularly in sectors like restaurants and entertainment.

C) By suggesting new usage occasions

Explanation: The text highlights that one way to increase the number of potential customers is by suggesting or developing new usage occasions for products, as illustrated by Arm & Hammer Baking Soda being used as a deodorizer.

B) Feature extraction

Explanation: Feature extraction is a vital step in data mining, where relevant features are identified and extracted from raw data to improve the performance of models.

C) It can predict values outside the range of 0 and 1

Explanation: Linear regression is not suitable for binary outcomes because it can produce predictions that fall outside the valid probability range of 0 to 1, leading to nonsensical results.

B) Intelligent Systems & Techniques

Explanation: The course IS4242 is specifically focused on Intelligent Systems & Techniques, indicating its relevance to the field of artificial intelligence and its applications.

B) Difficulty in managing interdependent parts

Explanation: As a venture reaches a critical stage, the complexity increases due to many interdependent and moving parts, making management difficult.

A) It is too high for the average response rate

Explanation: Given the low base rate of responses (0.17), the default cut-off value of 0.5 does not work effectively, necessitating a reevaluation of the threshold.

C) It measures the chance of observing data as extreme or more extreme under the null hypothesis

Explanation: Finding this probability helps to evaluate how likely it is to observe the test statistic or something more extreme if the null hypothesis is true, which is central to hypothesis testing.

B) Larger estimates are required to reject the null hypothesis

Explanation: A high standard error suggests that the estimates are less precise, requiring larger non-zero estimates to confidently reject the null hypothesis.

B) It can yield negative probabilities

Explanation: A significant limitation of linear regression is that it can produce negative probabilities, which are not valid in the context of probability theory, where all probabilities must lie between 0 and 1.

B) Intelligent Systems & Techniques

Explanation: The course IS4242 is specifically focused on Intelligent Systems & Techniques, indicating its relevance to the field of artificial intelligence and related technologies.

A) <6%

Explanation: Among the 40,000 companies founded and listed in the S&P Capital IQ database, less than 6% achieved over $10M in revenue by 2010, indicating the difficulty of scaling for startups.

D) Diagnostic Analytics

Explanation: The types of analytics mentioned are Descriptive, Predictive, and Prescriptive Analytics. Diagnostic Analytics is not listed among them.

B) Investing in advertising

Explanation: The text emphasizes that increasing acquisition expenditures can be achieved by investing in various advertising channels such as TV, radio, Google AdWords, and social media, which ultimately lead to website visits and sales.

B) Paying influencers on social media platforms

Explanation: Influencer marketing involves marketing companies paying influencers on platforms like Instagram and YouTube to generate credible word-of-mouth, which is a modern strategy for increasing awareness.

B) Using it as a deodorizer

Explanation: Arm & Hammer Baking Soda expanded its market by promoting its use as a deodorizer, demonstrating how new usage occasions can attract more customers.

B) By focusing on small businesses and then individuals

Explanation: Costco initially targeted small businesses but later expanded its warehouse club to individuals willing to pay an annual fee, illustrating a strategy to increase market size.

B) The probability of a positive outcome given certain predictors

Explanation: The notation P(Response = Yes | predictor variables) represents the conditional probability of a positive response (Yes) based on the values of the predictor variables, which is crucial in classification tasks.

C) It reduces the acquisition rate

Explanation: The text states that as the market expands, the acquisition rate tends to decrease, indicating that a broader market can lead to challenges in acquiring new customers.

B) Focusing on one customer at a time

Explanation: One-to-one marketing is characterized by its focus on individual customers, tailoring marketing efforts to meet the specific needs and preferences of each person.

C) Reducing operational costs

Explanation: Increasing acquisition expenditures focuses on generating awareness and attracting new customers, rather than reducing operational costs, which is not a direct benefit of this strategy.

B) They will drop significantly and lead to profits from retained customers

Explanation: Companies often believe that by reducing acquisition costs, they can earn profits from retained customers, highlighting a common misconception in business strategy.

C) It reduced by 250%

Explanation: Between 2000 and 2010, the cost of goods sold at the average S&P 500 company decreased significantly by 250%, indicating a major shift in operational efficiency.

B) Get the distribution of the t-statistic assuming the null hypothesis is true

Explanation: The first step in calculating a p-value involves obtaining the distribution of the t-statistic under the assumption that the null hypothesis is true.

B) Consumer confusion about brand positioning

Explanation: Increasing market size can lead to consumer confusion regarding the brand's positioning, as seen with BMW's expansion into lower-end offerings, which risks diluting its luxury image.

B) The variability of the coefficient estimate

Explanation: The variance of Dβ2 indicates how much the coefficient estimate can vary, which is crucial for understanding the reliability of the estimate in regression analysis.

B) 0 to 1

Explanation: Probabilities must always lie within the range of 0 to 1, making it essential to use appropriate models that respect this constraint when predicting outcomes.

B) Based on their demographics or purchase behavior

Explanation: Grouping customers according to demographics or purchase behavior allows businesses to identify patterns and similarities, which can inform strategies for acquiring new customers.

B) Model it as a linear regression problem

Explanation: The better approach suggested is to model the relationship as a linear regression problem, which allows for more accurate predictions of house prices based on attributes.

B) Regression techniques

Explanation: Regression techniques are highlighted as useful methods for identifying significant attributes that correlate with EVC, thereby facilitating the development of customized pricing.

B) By generating awareness

Explanation: Increasing acquisition expenditures helps in customer acquisition primarily by investing to generate awareness, which is crucial for attracting new customers to the company.

B) Investing in lead products

Explanation: Investing in lead products is a strategy to draw consumers to the company, making it a key aspect of increasing acquisition expenditures.

B) Consumer Lifetime Value

Explanation: The class IS4242 primarily focuses on Consumer Lifetime Value, which is a key concept in understanding the long-term value of customers to a business.

B) The response rate of a predictive model over the average response rate

Explanation: Lift calculates how much better a predictive model performs compared to the average response rate in the data, making it a crucial metric for evaluating model effectiveness.

B) Based on the probability of responding, ordered from highest to lowest

Explanation: Deciles are calculated by ordering the probabilities of responding from highest to lowest, allowing for a structured analysis of response rates across different segments.

B) The point that maximizes profit

Explanation: The optimal threshold for classification in the context of marketing campaigns is the point that maximizes profit, balancing true positives and false positives to achieve the best financial outcome.

B) Pricing strategies

Explanation: The title 'Techniques For Pricing' indicates that the document focuses on various strategies and methods related to pricing products or services.

C) They are twice as likely to respond compared to the average customer

Explanation: A lift value of 2 signifies that the customers in that segment are twice as likely to respond compared to the average customer, making it a critical threshold for targeting.

B) Almost none can drive down acquisition costs to a profitable level

Explanation: The reality is that very few companies manage to reduce their acquisition costs to a level that allows them to make a profit, which contradicts their initial beliefs.

C) They will go out of business or operate in small niches

Explanation: Promising ventures that cannot afford to burn through working capital face the risk of going out of business or being forced to operate in small niches.

B) P < 0.1 or 0.05

Explanation: In research and practice, p-values of less than 0.1 or 0.05 are commonly used as thresholds to determine statistical significance.

C) The initial query guiding the analysis

Explanation: The 'Business Question' serves as the foundational query that directs the entire analysis process in intelligent systems, ensuring that the analysis is relevant to business needs.

B) Smaller non-zero estimates may be sufficient

Explanation: A low standard error indicates that the estimates are more precise, meaning that even smaller non-zero estimates may be sufficient to reject the null hypothesis.

C) To clean and prepare data for analysis

Explanation: Data preprocessing is crucial in data mining as it involves cleaning and preparing the data, ensuring that it is suitable for further analysis and modeling.

B) Market trends and customer behavior

Explanation: Business questions in data mining are designed to uncover insights related to market trends and customer behavior, guiding decision-making processes.

B) Cross-validation techniques

Explanation: The text indicates that there will be a discussion about cross-validation techniques in the future, which are important for selecting the parameter C in soft-margin SVM.

B) To shrink coefficient estimates to zero

Explanation: LASSO (Least Absolute Shrinkage and Selection Operator) is designed to shrink coefficient estimates towards zero, which is particularly useful for feature selection by effectively removing variables with zero coefficients from the model.

B) Check the covariance or correlation

Explanation: The initial approach involves checking the covariance or correlation between each variable (house attributes) and the price to identify relationships.

C) September 10, 11:59 PM

Explanation: The assignment is specifically due on September 10 at 11:59 PM, which is crucial for students to note for timely submission.

D) Penalty for late submission

Explanation: The announcement clearly states that there will be a penalty for late submission, emphasizing the importance of starting the assignment early.

C) Sensitivity

Explanation: Recall is synonymous with sensitivity, measuring the proportion of actual positives that are correctly identified.

A) Consumer Acquisition

Explanation: Consumer Acquisition is explicitly mentioned as a key topic in the IS4242 class, indicating its importance in the curriculum.

B) Highly concentrated market where the winner takes all

Explanation: In a hyper-competitive market, the dynamics are such that a few companies dominate, and the one that performs best captures the majority of the market share, emphasizing the need for intelligent systems to stay competitive.

B) It is difficult to make predictions

Explanation: While looking at the distribution of predictors can help identify important variables, it poses a challenge in making accurate predictions, indicating a limitation of this approach.

C) 0.17

Explanation: The average response rate in the test data is specifically mentioned as 0.17, indicating a low base rate of responses which is crucial for determining the optimal threshold.

B) It can sink a business

Explanation: The cost of customer acquisition can be a major risk for businesses, especially in the digital landscape, where high costs can lead to financial difficulties.

B) Adding features iteratively to identify the best subset

Explanation: The simple method for feature selection in linear regression involves adding features iteratively, which helps in identifying the best subset of features that contribute to the model's performance.

C) It hurt the brand image by making Cadillac seem non-luxury

Explanation: The introduction of the Cimarron, a version of the Chevrolet Cavalier, negatively affected Cadillac's brand image, leading consumers to perceive it as a non-luxury vehicle.

C) 20 purchases

Explanation: A customer needed to make 20 purchases to break even, given the acquisition cost and profit margin, which illustrates the challenges in customer acquisition for Pets.com.

B) argmax α σ j α j - 1/2 σ j,k α j α k y j y k (φ(X j ) ∙ φ(X k ))

Explanation: The objective function is defined as argmax α σ j α j - 1/2 σ j,k α j α k y j y k (φ(X j ) ∙ φ(X k )), which aims to maximize the value of alpha in the transformed space.

B) K(X j , X k ) = (1 + X j ∙ X k )^d

Explanation: The polynomial kernel is defined as K(X j , X k ) = (1 + X j ∙ X k )^d, which is a specific form of kernel function used in SVM.

C) Identify customer preferences and behaviors

Explanation: Classification techniques are used to analyze customer data, allowing businesses to identify preferences and behaviors, which can inform targeted marketing efforts.

B) Ineffective marketing strategies

Explanation: Poor performance in identifying customers who will respond can lead to ineffective marketing strategies, as businesses may not target the right audience effectively.

B) Identifying attributes that correlate with EVC

Explanation: The main challenge in obtaining customized pricing is finding the attributes that correlate with the Economic Value to the Customer (EVC), which is essential for determining appropriate pricing strategies.

B) To measure and solve problems

Explanation: Companies need Intelligent Systems because data helps them measure performance and identify problems, which is essential for effective decision-making and problem-solving.

B) Data-driven products

Explanation: The data revolution has led to the creation of efficient processes and data-driven products, exemplified by companies like Tesla and Apple that integrate software with unique hardware.

B) Tastes

Explanation: The economic value of a product varies greatly among consumers due to factors such as individual tastes, which significantly influence how much value different consumers place on the same product.

A) True Positive

Explanation: In binary classification, TP stands for True Positive, which refers to the instances where the model correctly predicts the positive class.

C) Financial performance of the model

Explanation: The optimal threshold value should be based on the financial performance of the model, which includes metrics like lift value and Youden’s index, rather than arbitrary cut-off values.

C) To evaluate financial performance

Explanation: Lift value is used to assess the financial performance of the classification model, helping to set an optimal threshold that maximizes the model's effectiveness.

B) To generate website visits and sales

Explanation: The text indicates that by investing in advertising, the goal is to lead to website visits, which ultimately results in increased sales, highlighting the importance of acquisition expenditures.

B) It is a tuning parameter chosen via cross-validation

Explanation: The parameter C in soft-margin SVM is treated as a tuning parameter that is generally chosen via cross-validation, allowing for optimal model performance.

C) Reducing the acquisition rate

Explanation: Expanding the market to reach new segments can lead to a reduction in the acquisition rate, which is a significant risk that businesses must consider when planning market expansion.

C) Segmenting existing customers

Explanation: The classification techniques are primarily aimed at segmenting existing customers to better target them with marketing strategies, enhancing customer engagement and retention.

C) Survey on potential to capture value

Explanation: The text refers to a survey that assesses the potential for a firm to capture the value it creates, indicating a strategic approach to pricing.

D) Replace with random values

Explanation: Replacing missing values with random values is not a standard method for handling missing data, as it can introduce bias and inaccuracies. The other options are recognized methods.

E) The product's price point

Explanation: While models consider when products are likely to be bought and customer response to offers, they do not specifically focus on the product's price point as a primary factor.

D) 78 students

Explanation: The announcement mentions that there are 78 students in total, which allows for the formation of 26 groups of three members each.

B) A product of weights and labels

Explanation: The term σjαjyk represents a product involving the weights (σj and αj) and the labels (yk), which is part of the constraint that must equal zero.

B) Partitioning clustering

Explanation: K-means is a type of partitioning clustering method that divides the dataset into K distinct clusters based on distance to the centroid of each cluster.

C) Logistic Regression

Explanation: Logistic Regression is explicitly mentioned as one of the classification models, alongside Support Vector Machines (SVM), highlighting its relevance in classification tasks.

C) SG&A (Selling, General and Administrative expenses)

Explanation: Despite the significant reduction in the cost of goods sold, SG&A as a percentage of revenue did not change, highlighting a disconnect in cost management.

B) Predictive Analytics

Explanation: Predictive Analytics is specifically designed to forecast future outcomes by analyzing historical data, making it a vital component of intelligent systems.

B) Lack of exclusivity and damage to its image

Explanation: By expanding its brand reach with lower-end offerings, BMW risks losing its exclusivity and potentially damaging its luxury image, which is crucial for its brand identity.

B) Cadillac's Cimarron

Explanation: The Cadillac Cimarron serves as a prime example of brand dilution in the luxury market, as it was perceived as a less luxurious offering, damaging Cadillac's overall brand image.

C) To handle non-linear relationships in data

Explanation: Non-linear kernels in SVM are used to handle non-linear relationships in data, allowing the model to classify data that is not linearly separable.

B) It treats all customers as one group

Explanation: Mass marketing is defined by its approach of treating all customers as a single group, aiming to reach a broad audience without differentiation.

C) Higher product prices

Explanation: While classification techniques can lead to improved insights and efficiency, they do not inherently result in higher product prices; rather, they focus on understanding and targeting customers more effectively.

B) When data is not linearly separable

Explanation: SVMs with non-linear kernels are beneficial when the data cannot be separated by a straight line, allowing for more complex decision boundaries.

B) Capturing the value created by the firm

Explanation: Value-oriented pricing emphasizes the importance of capturing the value that a firm creates within its industry, rather than solely focusing on costs or competition.

C) The benefit, not the feature

Explanation: The emphasis is on identifying the benefit that the product provides, distinguishing it from merely listing its features, which is crucial for effective value communication.

B) TP / (TP + FP)

Explanation: Precision is calculated as the ratio of true positives (TP) to the sum of true positives and false positives (FP), indicating the accuracy of positive predictions.

D) 24%

Explanation: Firms that adopted value-oriented pricing strategies achieved 24% higher profits than their industry peers, indicating the effectiveness of this pricing approach.

B) Marketing and communication efforts

Explanation: Retaining customers involves costs such as mailings, phone calls, and targeted advertising on platforms like Google or Facebook, which can be significant.

B) A customer who is worth pursuing

Explanation: A target customer is defined as one who is worth pursuing, meaning they are expected to generate more revenue than the costs associated with sales and support.

B) The significance of a coefficient

Explanation: The T-statistic is used in hypothesis testing to determine whether a coefficient is significantly different from zero, which helps assess the impact of the independent variable in regression analysis.

A) False Negative

Explanation: FN stands for False Negative, which indicates the instances where the model incorrectly predicts the negative class when the true class is positive.

C) $100

Explanation: The average customer spent $100 per purchase on Pets.com, which is crucial for understanding the revenue potential and profitability of acquiring new customers.

B) Some errors allowed

Explanation: The soft-margin SVM allows for some errors by slightly changing the constraints, making it more flexible in handling noisy data compared to a hard-margin SVM.

B) It allows for the possibility of misclassification

Explanation: The soft-margin SVM formulation is characterized by its allowance for misclassification, making it suitable for datasets with noise.

B) There is no relationship between X and Y

Explanation: The null hypothesis (H0) asserts that there is no relationship between the independent variable (X) and the dependent variable (Y), represented mathematically as β1 = 0.

C) The coefficients are statistically significant

Explanation: A significant p-value for coefficients suggests that the coefficients are statistically significant, indicating that they have a meaningful impact on the dependent variable in the model.

B) It is not a good model for this purpose

Explanation: The statement indicates that the model is not good at identifying the variables that make an impact on sales, suggesting a significant flaw in its predictive capabilities.

B) Increased profitability

Explanation: By effectively capturing the value created, firms can enhance their profitability, making value-oriented pricing a strategic advantage.

B) No labeled training data

Explanation: Unsupervised learning is defined by the absence of labeled training data, allowing algorithms to discover patterns and structures within the data without prior labels.

C) Outliers

Explanation: Outliers are defined as extreme values that stand out from the rest of the data, which can skew results and affect the accuracy of data analysis.

B) To clean and prepare data for analysis

Explanation: Data preprocessing is essential in business intelligence as it involves cleaning and preparing data, ensuring that it is suitable for analysis and decision-making processes.

B) Training data

Explanation: Supervised learning relies on training data, which consists of input-output pairs (X, y) that help the model learn the relationship between the inputs and the corresponding outputs.

C) Data visualization

Explanation: Data visualization is not a step in data preprocessing; rather, it is a technique used after data has been processed to present the data in a visual format for easier interpretation.

C) $11.8 million

Explanation: Pets.com spent $11.8 million on advertising during its first fiscal year, which indicates the heavy investment in marketing to attract customers.

C) Foodies who enjoy higher quality products

Explanation: Whole Foods began as an organic grocery store but expanded its market by targeting 'foodies' who appreciate higher quality products, showcasing a strategy to attract new customer segments.

C) Less returns on invested capital

Explanation: A lower acquisition rate implies that there will be less return on the invested capital, which can lead to negative profits, highlighting the financial risks of market expansion.

C) It is complex but not effective in identifying impactful variables

Explanation: The model is described as complex, yet it is noted that it does not effectively identify the variables that significantly impact sales, highlighting a key limitation.

B) More variables can have non-zero coefficients

Explanation: A smaller λ value in LASSO regression reduces the regularization effect, allowing more variables to retain non-zero coefficients, which can lead to a more complex model.

B) K(X j , X k ) = e^{-γ |X j - X k |^2}

Explanation: The Radial Basis Function kernel is expressed as K(X j , X k ) = e^{-γ |X j - X k |^2}, which is commonly used in SVM for non-linear data.

B) Cost of production

Explanation: Value-oriented pricing does not primarily focus on the cost of production; instead, it centers on the value perceived by customers and the value created by the firm.

B) Comparing Delta Airlines to American Airlines

Explanation: The best alternative for Delta Airlines is American Airlines, as they are direct competitors in the airline industry, unlike trains or cars, which are not comparable in the same context.

C) Response (binary variable)

Explanation: The outcome in classification tasks is specified as a response variable that is binary, indicating that it can take on two possible values, which is a common characteristic in classification problems.

B) Difficulty in measuring consumer value

Explanation: One of the main challenges of value-oriented pricing is accurately assessing the economic value of a product to consumers, which can complicate the pricing strategy.

A) Cost of Goods Sold

Explanation: COGS stands for Cost of Goods Sold, which is a crucial component in determining pricing strategies and calculating profit margins.

C) Data Visualization

Explanation: Data visualization is not a focus area in data cleaning; rather, it is a method used to represent data visually after it has been cleaned and processed.

C) To increase the volume or value of existing purchases

Explanation: Upselling is aimed at encouraging customers to buy more or higher-value products they are already purchasing, thereby increasing overall sales.

B) Characteristics or attributes of the data

Explanation: In data mining, 'features' refer to the characteristics or attributes of the data that are used to build models and make predictions.

C) Broader markets lead to lower acquisition rates

Explanation: The text clearly states that the broader the market is, the lower the acquisition rate, indicating a negative correlation between market breadth and acquisition efficiency.

B) It focuses on selected groups of customers

Explanation: Target marketing lies between mass marketing and one-to-one marketing, as it specifically targets selected groups or market segments rather than treating all customers as one.

C) It drives the development of intelligent systems

Explanation: The text suggests that companies have always pursued data because it is essential for the creation and enhancement of intelligent systems, which are integral to modern business strategies.

B) To target these groups accordingly

Explanation: Designing products tailored to specific customer groups enables businesses to meet the unique needs and preferences of those segments, thereby enhancing customer satisfaction and acquisition.

B) Surveys or historical data of consumers and product attributes

Explanation: Surveys or historical data of consumers and product attributes are crucial for identifying significant attributes that correlate with EVC, aiding in the development of customized pricing strategies.

B) Missing Values

Explanation: Missing values refer to incomplete data entries that can significantly affect data analysis and require specific techniques for handling during the data cleaning process.

B) Selling a customer a higher volume or upgraded version of a product

Explanation: Upselling refers to encouraging customers to purchase more or upgraded versions of products they are already buying, as shown in the example of increasing life insurance coverage.

C) αj must be greater than or equal to 0

Explanation: The constraint states that αj must be non-negative (αj ≥ 0) for all j, which is a common requirement in optimization problems.

B) By measuring how much value these create

Explanation: It is important to measure the value created by differentiation to understand its impact on the product's market position and pricing strategy.

B) K-Means

Explanation: K-Means is identified as a method of Unsupervised Learning, highlighting its relevance in clustering data without prior labels.

C) To normalize the range of data values

Explanation: Scaling is used in data transformation to normalize the range of data values, ensuring that different features contribute equally to the analysis.

C) Lagrange method

Explanation: The maximization problem in SVM is solved using the Lagrange method, which is a technique for finding the local maxima and minima of a function subject to equality constraints.

B) There is a non-zero relationship between X and Y

Explanation: The alternate hypothesis (H1) suggests that there is a non-zero relationship between X and Y, which is represented as β1 ≠ 0.

B) A larger λ increases regularization, pushing more coefficients towards zero

Explanation: In LASSO regression, a larger value of the regularization parameter λ increases the amount of regularization applied, which leads to more coefficients being pushed towards zero, thereby enhancing feature selection.

B) It has experienced many ups and downs

Explanation: The statement highlights that the evolution of AI has not been straightforward, indicating a history of challenges and fluctuations in its development.

B) Learn as much as we can about current customers

Explanation: The initial step in acquiring new customers involves gathering detailed information about current customers and identifying their similarities, which helps in targeting potential new customers effectively.

B) Selling different products to the same customer

Explanation: Cross-selling involves offering different products to existing customers, as illustrated by the example of selling Quicken to TurboTax users.

B) To maximize the value of α

Explanation: The objective function is designed to maximize α, which is indicated by the notation 'argmax α' in the provided expression.

C) Errors

Explanation: Errors in data refer to inaccuracies or inconsistencies that can arise from various sources, such as data entry mistakes, and must be corrected to ensure data quality.

C) Reducing prices

Explanation: The text focuses on cross-selling and upselling as methods to increase marginal revenues, rather than reducing prices, which is not mentioned as a strategy.

B) To bring variables to the same scale or range

Explanation: Scaling is a process in data transformation that aims to bring different variables to the same scale or range, which is essential for many machine learning algorithms to function effectively.

C) To clean and prepare data for analysis

Explanation: Data preprocessing is essential in data mining as it involves cleaning and preparing data to ensure that it is suitable for analysis, which enhances the quality of the results.

B) [0, 1]

Explanation: Normalization typically transforms data to a range of [0, 1], making it easier to compare and analyze different variables on a common scale.

C) To improve marketing strategies

Explanation: The goal of predicting customer responses is to improve marketing strategies by identifying which customers are likely to respond positively, thereby optimizing outreach efforts.

B) Numerical or ordinal

Explanation: In supervised learning, regression is used when the output (y) is numerical or ordinal, allowing for the prediction of continuous values.

B) Negative profits

Explanation: The text indicates that a lower acquisition rate resulting from market expansion can lead to negative profits, emphasizing the potential financial drawbacks of such strategies.

C) Data

Explanation: The text emphasizes that despite the ups and downs in the evolution of AI, companies have consistently focused on acquiring and utilizing data, which is crucial for developing intelligent systems.

B) It increases customer expenditures with the firm

Explanation: Target marketing involves direct marketing to customers who are most likely to buy, which can lead to increased customer expenditures and higher sales for the firm.

B) Through cross-validation technique

Explanation: The optimal value of the regularization parameter λ in LASSO regression can be obtained through cross-validation, which helps in selecting the best λ that minimizes prediction error.

B) It does not help in prediction

Explanation: While correlation indicates the direction of the relationship, it does not provide predictive power, which is essential for modeling house prices effectively.

B) Any speakers, headphones, or earbuds

Explanation: Casual listeners would be targeted with a range of accessible audio products such as speakers, headphones, or earbuds, which cater to their general listening needs.

B) Economic value of the product to consumers (EVC)

Explanation: Value-oriented pricing emphasizes understanding and leveraging the economic value that a product provides to consumers, which is crucial for setting an effective price.

B) New customers are more expensive to acquire

Explanation: It is stated that acquiring a new customer can cost five to seven times more than retaining an existing one, making it easier to extract profit from current customers.

D) Maximize 2 | |W| |

Explanation: The primary objective in the SVM optimization problem is to maximize 2 | |W| |, which can also be equivalently expressed as minimizing | |W| | or | |W| |^2 for mathematical convenience.

B) Regression and Classification

Explanation: The techniques highlighted for targeting current customers include Regression and Classification, which are essential for analyzing customer behavior and predicting future purchases.

A) True Positive

Explanation: TP stands for True Positive, which refers to the instances where the model correctly predicts the positive class.

B) To understand how prices vary with independent variables

Explanation: The main goal of modeling house prices is to understand how various independent variables affect the prices, which can help management make informed decisions.

B) Removing inaccuracies and inconsistencies

Explanation: Data cleaning focuses on identifying and correcting inaccuracies and inconsistencies in the data, which is crucial for ensuring the quality of the data used in analysis.

C) p-values of the t-statistic

Explanation: Hypothesis testing in linear regression is conducted using p-values derived from the t-statistic, which helps determine the significance of the relationship between the variables.

B) K(X j , X k ) = φ(X j ) ∙ φ(X k )

Explanation: The kernel function is defined as K(X j , X k ) = φ(X j ) ∙ φ(X k ), which computes the similarity in the transformed space.

B) They perform poorly on class 1

Explanation: The text highlights that both Logistic Regression and SVM are performing poorly specifically on class 1, indicating a challenge in accurately identifying certain customer responses.

B) Change the thresholds of the classifier

Explanation: The text suggests that one way to address the poor performance on class 1 is to change the thresholds of the classifier, which can help improve the identification of customers who will respond.

C) Data privacy concerns

Explanation: One of the significant challenges in using classification techniques is ensuring data privacy and compliance with regulations, which can affect how customer data is collected and used.

B) To improve marginal revenues

Explanation: The text emphasizes that by targeting current customers, firms can enhance their marginal revenues through strategies like cross-selling and upselling.

B) It offers insights into value-oriented pricing

Explanation: The source cited is relevant for understanding the principles and potential of value-oriented pricing strategies in capturing the value created by firms.

C) Potential sources of differentiation value

Explanation: Identifying potential sources of differentiation value is essential as it follows from the benefits provided by the product and helps in establishing a unique market position.

B) Tesla

Explanation: Tesla is cited as an example of a company that develops cars by integrating software with unique hardware, showcasing the application of Intelligent Systems in product development.

B) Classification

Explanation: The task specified for the model is classification, which involves predicting the probability that a customer will respond positively.

B) Capture a portion of the economic value

Explanation: The goal of value-oriented pricing is to capture a portion of the economic value that the product provides to consumers, allowing the firm to benefit financially from its offerings.

B) To optimize inventory management

Explanation: Understanding when a product is likely to be bought helps businesses optimize their inventory management, ensuring that products are available when customers are ready to purchase.

B) To improve data accuracy and quality

Explanation: The primary goal of data cleaning is to enhance the accuracy and quality of data, ensuring that it is reliable for analysis and decision-making.

B) To understand the relation between house attributes and price

Explanation: The main task is to understand how different house attributes relate to the price, which is crucial for making informed decisions in real estate.

B) Neural networks

Explanation: Neural networks are a popular method used in classification techniques to analyze complex customer data and improve targeting strategies.

A) Casual listeners and music lovers

Explanation: The example illustrates how an audio equipment company can categorize its customers into groups like casual listeners and music lovers, each requiring different types of products.

B) Remove columns or refrain from using them in the model

Explanation: One common approach to handle missing values is to remove columns that contain them or to refrain from using those columns in the model, especially if they are not critical to the analysis.

C) Prices

Explanation: When obtaining EVC is challenging, prices can be used as a proxy, allowing businesses to estimate the economic value based on existing pricing structures.

C) Three members

Explanation: Each group is allowed to have three members, which is specified in the announcement to facilitate group formation.

B) K-means

Explanation: K-means is a popular clustering method used in unsupervised learning to partition data into distinct groups based on similarity.

C) y_i W ∙ X_i + b ≥ 1

Explanation: The constraints in the SVM optimization problem are defined as y_i W ∙ X_i + b ≥ 1, ensuring that the data points are correctly classified with a margin.

B) Price - COGS

Explanation: The firm's incentive to sell is calculated as Price minus COGS, indicating the profit made on each unit sold.

B) They can distort statistical analyses

Explanation: Outliers can significantly distort statistical analyses and model performance, leading to misleading results if not properly addressed during data cleaning.

B) To launch a gold membership offering a 20% discount

Explanation: The superstore is planning to introduce a gold membership that provides a 20% discount on all purchases, targeting existing customers as part of their year-end sale strategy.

C) It increases market share and revenue

Explanation: Finding new customers is crucial in intelligent systems as it directly contributes to increasing market share and revenue, which are essential for business growth.

B) To prepare data for analytics

Explanation: Data preprocessing is essential for preparing real-world data for analytics, as it addresses issues of dirtiness, misalignment, and complexity.

B) Classification performance at all thresholds

Explanation: The ROC curve provides a comprehensive view of classification performance across all possible thresholds, allowing for the evaluation of true positive and false positive rates.

B) What product the customer is likely to buy next

Explanation: The models for targeting current customers are designed to predict which product a customer is likely to purchase next, helping businesses tailor their marketing strategies effectively.

B) To discover structure in data

Explanation: The main objective of unsupervised learning is to uncover hidden patterns or structures in data, which can be useful for various applications such as clustering and segmentation.

B) 2 * P * R / (P + R)

Explanation: The F1-Measure is calculated using the harmonic mean of precision (P) and recall (R), making it particularly useful for imbalanced datasets.

B) Using the mean, median, or mode

Explanation: A common method for imputing missing values is to use statistical measures such as the mean, median, or mode, which helps to maintain the dataset's overall distribution.

C) Acquiring new customers can cost five to seven times more

Explanation: The text explicitly states that acquiring a new customer can be significantly more expensive than retaining an existing one, highlighting the financial advantage of focusing on current customers.

B) Intelligent Systems & Techniques

Explanation: The course IS4242 is specifically centered around Intelligent Systems & Techniques, indicating its focus on the application of intelligent systems in various contexts.

B) A technique that creates a tree-like structure of clusters

Explanation: Hierarchical clustering is a method that builds a hierarchy of clusters, often represented as a dendrogram, allowing for the exploration of data at various levels of granularity.

B) Applying a predetermined markup to the cost of production

Explanation: Cost-plus pricing involves calculating the total cost of production and then adding a predetermined markup to determine the final price, making it a straightforward pricing strategy.

B) Retaining valuable customers

Explanation: The text emphasizes the importance of targeting and retaining valuable customers, suggesting that this is a more profitable strategy than acquiring new ones.

C) Charge lower than the maximum price

Explanation: It is suggested to charge a price lower than the maximum to foster strong customer relations, as this can enhance customer satisfaction and loyalty.

A) The proportion of true positives among all actual positives

Explanation: Sensitivity is defined as the ratio of true positives (TP) to the sum of true positives and false negatives (TP + FN), indicating how effectively a test identifies actual positive cases.

C) September 10, 11:59 PM

Explanation: The announcement clearly states that Programming Assignment 1 is due on September 10 at 11:59 PM, indicating the deadline for submission.

B) Data distribution analysis

Explanation: Outliers can be identified using data distribution analysis, which helps in recognizing extreme or unrealistic values within the dataset.

B) Remove observations (if not many)

Explanation: If there are only a few missing observations, it is often acceptable to remove those specific observations from the dataset to maintain the integrity of the analysis.

B) Look at the distribution of predictors for each class

Explanation: A simple way to identify important variables is to examine the distribution of predictors for each class, which helps in recognizing variables that have significantly different distributions.

C) Created a hyper-turbulent and hyper-competitive market

Explanation: Digital transformation has led to a hyper-turbulent and hyper-competitive market across almost every industry, emphasizing the need for companies to adapt and innovate.

B) To target customers through phone calls

Explanation: The management intends to use the model to specifically target customers through phone calls, indicating a direct application of the model's predictions.

D) Competitive pricing

Explanation: While value-oriented pricing focuses on the economic value to consumers and capturing that value, it does not primarily emphasize competitive pricing, which is more about matching or beating competitors' prices.

B) Economic Value - Price

Explanation: The customer's incentive to purchase is determined by the Economic Value minus the Price, reflecting the perceived benefit of the product relative to its cost.

B) It does not account for consumer differences

Explanation: Uniform pricing is sub-optimal because it fails to consider the diverse economic values that different consumers assign to a product based on their unique characteristics and usage patterns.

C) Existing customers only

Explanation: The gold membership discount is specifically valid for existing customers, indicating a targeted approach to retain current clientele during the year-end sale.

B) To reduce overfitting and improve model accuracy

Explanation: The primary purpose of LASSO is to reduce overfitting by penalizing the absolute size of the coefficients, which helps in improving the model's accuracy and interpretability by selecting only the most significant predictors.

B) AI-first companies

Explanation: Companies need to adopt an AI-first approach to effectively leverage data for strategic decision-making, which is essential for building a competitive advantage in a hyper-competitive environment.

B) It combines data from different sources for a comprehensive view

Explanation: Data integration is crucial as it allows for the combination of data from various sources, providing a more comprehensive view that enhances the quality of analysis and decision-making.

E) Product packaging

Explanation: While product packaging can affect consumer perception, it is not listed as a direct factor influencing the economic value of a product, which is primarily determined by tastes, nature of use, and intensity of use.

B) Efficient search functionality

Explanation: Implementing efficient search functionality is a straightforward method to assist customers in finding quality products they are interested in, enhancing their shopping experience.

B) Highly volatile market with difficulty in maintaining a sustainable advantage

Explanation: A hyper-turbulent market is characterized by rapid changes and volatility, making it challenging for companies to maintain a competitive edge over time, thus necessitating the use of intelligent systems.

A) False Negative

Explanation: FN stands for False Negative, indicating the instances where the model incorrectly predicts the negative class when the actual class is positive.

B) Selecting important variables from raw data

Explanation: Feature extraction is the process of identifying and selecting the most relevant variables from raw data, which helps in improving the efficiency and accuracy of data mining tasks.

C) Whether the customer accepted the offer in the last campaign

Explanation: The outcome variable, or response, is defined as 1 if the customer accepted the offer in the last campaign and 0 otherwise, indicating its role as a target variable in predictive modeling.

B) Customers got tired of receiving discount offers

Explanation: Groupon's marketing strategy led to customer fatigue regarding constant discount offers, indicating a need for a more effective approach to customer engagement.

A) Data Mining

Explanation: Data Mining is a key technique used in finding new customers, as it allows businesses to analyze large sets of data to identify potential customer segments and trends.

C) Principal Component Analysis

Explanation: Principal Component Analysis (PCA) is highlighted as a technique for Dimensionality Reduction, showcasing its utility in transforming data into a lower-dimensional space.

B) Expected net present value of future profit contributions by a customer

Explanation: LTV is defined as the expected net present value of future profit contributions from a customer after acquisition, highlighting its importance in understanding customer profitability over time.

B) Model Analyze/Explore/Predict

Explanation: A common task in data mining is to analyze, explore, and predict outcomes based on the data, which is essential for making informed business decisions.

C) Discount rate

Explanation: In the LTV formula, δ represents the discount rate, which is used to calculate the present value of future cash flows from a customer.

B) OverallQual

Explanation: 'OverallQual' rates the overall material and finish of the house on a scale from 1 to 10, providing a qualitative measure of the property's construction quality.

B) Collecting data and extracting meaningful insights

Explanation: An AI-first company focuses on leveraging data to extract insights that inform decision-making, which is essential for modern business strategies.

A) Converting continuous data into categorical data

Explanation: Discretization is the process of converting continuous data into categorical data, which can help in simplifying models and making them easier to interpret.

B) Their sales revenue exceeds costs of sales and support

Explanation: A profitable customer is characterized by having sales revenue that exceeds the costs associated with sales and support, making them valuable to the business.

B) Agglomerative Clustering

Explanation: Agglomerative Clustering is noted as another technique under Unsupervised Learning, emphasizing its role in grouping data points based on similarity.

A) Combining multiple data points into a single summary value

Explanation: Aggregation involves combining multiple data points into a single summary value, which helps in simplifying the dataset and making it easier to analyze.

B) To simplify data while retaining important information

Explanation: Dimensionality Reduction aims to reduce the number of features in a dataset while preserving essential information, making it easier to analyze and visualize.

D) 25%

Explanation: An example given in the content indicates that a markup of 25% is often added to the total cost of production in cost-plus pricing, illustrating a typical approach in this pricing strategy.

B) To assess the profitability of a customer over time

Explanation: The primary purpose of calculating Customer Lifetime Value (LTV) is to assess the long-term profitability of a customer, helping businesses make informed decisions about targeting and resource allocation.

C) Making a predictive model to identify potential customers

Explanation: The management aims to create a predictive model to identify customers who are likely to purchase the gold membership, which would help in reducing the overall campaign costs.

C) It applies L1 regularization

Explanation: LASSO regression is characterized by its use of L1 regularization, which adds a penalty equal to the absolute value of the magnitude of coefficients, allowing for both shrinkage and variable selection.

C) Price should be lower than economic value for better sales

Explanation: For better sales and customer satisfaction, the price should ideally be lower than the economic value perceived by the consumer, allowing for a favorable purchasing decision.

B) To know what is happening

Explanation: Companies aim to understand their current situation, which is a fundamental aspect of utilizing intelligent systems for business decision-making.

C) The odds of the outcome occurring

Explanation: The equation log(P(Y)/(1-P(Y))) represents the log-odds of the outcome occurring in logistic regression, which is a key concept in understanding how logistic regression models probabilities.

B) To understand price dynamics and optimize pricing

Explanation: The expected outcome is to gain insights into how various factors affect prices, allowing for optimized pricing strategies that can yield better financial returns.

B) Discount rate

Explanation: The variable δ in the LTV formula represents the discount rate, which is used to account for the time value of money in calculating the present value of future cash flows.

A) The relationship between input and output

Explanation: The learning function, denoted as y = g(X), represents the relationship between the input features (X) and the output (y) that the model aims to learn.

B) It helps in measuring performance

Explanation: Utilizing data allows companies to measure performance effectively, which is crucial for identifying issues and making informed decisions.

B) To understand price variations

Explanation: The management will use the model to comprehend how prices will vary with different independent variables, aiding in strategic decision-making.

A) Normalization

Explanation: Normalization is a common technique used in data transformation to adjust the values in the dataset to a common scale, which helps in improving the accuracy of analysis.

A) Inconsistent class labels

Explanation: Missing values can lead to issues such as inconsistent class labels, which can affect the integrity and usability of the dataset during analysis.

B) To measure the strength and direction of relationships between variables

Explanation: Coefficient estimates are used in statistical modeling to quantify the relationship between independent and dependent variables, indicating how changes in one variable affect another.

E) Customer's favorite color

Explanation: The data description lists various predictors related to customer characteristics and purchase behavior, but it does not include the customer's favorite color, making it a distractor.

C) Leaving revenue on the table

Explanation: Charging too little for a product can result in leaving potential revenue uncollected, which is a significant concern in pricing strategy.

C) They analyze data to identify potential customers

Explanation: Intelligent systems are instrumental in analyzing data to identify potential customers, enabling businesses to target their marketing efforts more effectively.

B) Profit = (Price - Cost) * Volume

Explanation: The correct formula for calculating profit is Profit = (Price - Cost) * Volume, which highlights the relationship between pricing, costs, and sales volume in determining profitability.

C) Experience

Explanation: The effectiveness and techniques of data preprocessing improve and develop with experience, highlighting the subjective nature of the process.

B) Model analyze/explore/predict

Explanation: A key task in data mining involves analyzing, exploring, and predicting outcomes using models, which helps in extracting valuable insights from data.

D) It increases exponentially

Explanation: The complexity of feature selection increases exponentially with the number of variables, making it more challenging to identify the best subset as the number of features grows.

B) They should be treated as missing values or errors

Explanation: Extreme or unrealistic values are often treated as missing values or errors during the data cleaning process to ensure data integrity.

B) A dot product of transformed features

Explanation: The notation φ(Xj) ∙ φ(Xk) indicates the dot product of the transformed features, which is a common operation in machine learning to measure similarity.

B) Membership offers and cross-selling/up-selling offers

Explanation: The models specifically assess how likely customers are to respond to membership offers and cross-selling or up-selling offers, which are crucial for maximizing customer value.

C) To ensure decisions are made based on facts rather than whims

Explanation: In a rapidly changing market, relying on data for decision-making is crucial to adapt and respond effectively, rather than making arbitrary choices that could lead to failure.

C) Support Vector Machines (SVM)

Explanation: Support Vector Machines (SVM) is mentioned as another classification model, indicating its importance in the context of classification tasks.

B) A question that guides the data mining process to achieve specific business objectives

Explanation: A business question in data mining is crucial as it directs the analysis and helps in achieving specific business goals, ensuring that the data mining efforts are aligned with organizational needs.

E) Predictive Learning

Explanation: Predictive Learning is not listed as a learning paradigm of data mining tasks. The recognized paradigms include Supervised, Unsupervised, Semi-supervised, Active Learning, and Reinforcement Learning.

B) Combining multiple data points into a single summary

Explanation: Aggregation involves combining multiple data points into a single summary, which helps in simplifying the dataset and making it easier to analyze.

A) EVC = Price of the alternative + value differential

Explanation: The formula for calculating Economic Value (EVC) is given as the price of the alternative plus the value differential, which is crucial for determining the economic viability of the server alternatives for the toy company.

A) It uses labeled data for training

Explanation: Supervised learning is characterized by the use of labeled data, where the model learns from input-output pairs to make predictions on new data.

A) To convert numerical data into categorical data

Explanation: Discretization is used to convert numerical data into categorical data by reducing the number of distinct values, making it easier to analyze and interpret.

B) Characteristics or properties used for analysis

Explanation: Features in data mining refer to the characteristics or properties of the data that are used for analysis, helping to identify patterns and make predictions.

C) 3.3%

Explanation: A 1% improvement in unit volume, while maintaining the same price, results in a 3.3% increase in operating profit, demonstrating the significant impact of volume on profitability.

C) 20%

Explanation: The gold membership offers a 20% discount on all purchases, which is a significant incentive for existing customers to participate in the year-end sale.

C) Mailings and targeted advertising

Explanation: The text mentions strategies such as mailings, phone calls, and targeted advertising on platforms like Google or Facebook as methods to target and retain valuable customers.

A) Least Absolute Shrinkage and Selection Operator

Explanation: LASSO stands for Least Absolute Shrinkage and Selection Operator, which is a regression analysis method that performs both variable selection and regularization to enhance the prediction accuracy and interpretability of the statistical model.

B) The average value of the variable

Explanation: In data transformation, the variable mean (𝜇) represents the average value of the variable, which is used in normalization and standardization processes.

B) Data Preprocessing

Explanation: Data preprocessing is a crucial initial step in developing an intelligent system, as it involves preparing and cleaning the data for further analysis and feature extraction.

B) Feature extraction

Explanation: Data preprocessing primarily involves feature extraction, which is crucial for preparing data for analysis and ensuring that the most relevant information is used in data mining tasks.

B) To combine multiple data points into a single summary

Explanation: Aggregation is a process in data transformation that combines multiple data points into a single summary, which can simplify analysis and interpretation.

D) Normalization

Explanation: Normalization is a process of scaling data, not a method of constructing new variables. The other options involve creating new variables through mathematical operations.

B) The model's output for a given input

Explanation: The prediction function g(Xi) represents the model's output for a specific input Xi, indicating how the model predicts the corresponding output based on the learned relationship.

B) R-squared

Explanation: The best subset of features is identified based on R-squared, which measures the proportion of variance in the dependent variable that can be explained by the independent variables.

C) They were unprofitable and did not come back

Explanation: Merchants discovered that customers attracted through Groupons often turned out to be unprofitable, as they did not return for future purchases.

B) True positive rate - false positive rate

Explanation: Youden’s index is calculated as the difference between the true positive rate (sensitivity) and the false positive rate (1 - specificity), providing a measure of classification effectiveness.

B) Unsupervised

Explanation: Unsupervised learning involves learning from unlabeled data, allowing the model to identify patterns and structures without predefined labels.

C) Amount spent on fish

Explanation: The amount spent on various products, including fish, is part of the customer purchase behavior attributes, which help in understanding spending patterns.

B) Revenue and cost from the consumer at time t

Explanation: R_t and C_t denote the revenue and cost associated with a consumer at time t, which are essential components in calculating the Customer Lifetime Value.

C) It is dirty, misaligned, and overly complex

Explanation: Real-world data is frequently described as dirty, misaligned, overly complex, and often inaccurate, making it unsuitable for immediate analytics without preprocessing.

B) For mathematical convenience

Explanation: The form of the SVM optimization problem is chosen for mathematical convenience, allowing for easier manipulation and solution of the optimization problem.

B) To ensure products are priced appropriately

Explanation: The text emphasizes that nothing is more important than ensuring products are priced appropriately, as incorrect pricing can lead to lost revenue or customer alienation.

B) To create binary variables from categorical data

Explanation: One-hot encoding is a technique used to convert categorical variables into a binary format, allowing them to be used in machine learning algorithms effectively.

B) Total Assets of a company is negative

Explanation: An example of an error in data cleaning is when the total assets of a company are recorded as negative, which is logically inconsistent and needs correction.

B) A question that requires data analysis to inform business decisions

Explanation: A business question in data mining is one that necessitates data analysis to derive insights that can guide business strategies and decisions.

C) To identify potential customers for outreach

Explanation: The expected outcome of the model's predictions is to identify potential customers who are likely to respond positively, facilitating targeted outreach efforts.

C) It has a limited ability to capture the customer’s willingness to pay

Explanation: A significant drawback of cost-plus pricing is that it may not effectively account for how much customers are willing to pay, which can result in lost revenue opportunities.

C) Higher costs

Explanation: Using intelligent systems typically leads to increased efficiency and improved targeting, rather than higher costs, making it a cost-effective approach for finding new customers.

B) Designing products to signal customer value through choice

Explanation: The strategy involves designing products in a way that allows customers to signal their perceived value (high or low) through their product choices, which can help in differentiating market segments.

B) It combines labeled and unlabeled data

Explanation: Semi-supervised learning combines both labeled and unlabeled data, leveraging the strengths of both to improve learning accuracy.

B) To purchase houses below their actual values and sell them at a higher price

Explanation: Surprise Housing's strategy involves buying houses at prices lower than their actual values and then flipping them for a profit, which is a common practice in real estate investment.

C) 12 billion products

Explanation: Amazon boasts a vast catalog of 12 billion products, highlighting its extensive inventory and the scale of its operations in matching demand and supply.

B) It enforces a balance between positive and negative classes

Explanation: The constraint σjαj yj = 0 ensures that the weighted sum of the labels (yj) is balanced, which is crucial for classification tasks.

B) Few suppliers have high demand while many have low demand

Explanation: Amazon's supplier landscape features a long tail, where only a few suppliers experience high demand, while many others have very low demand, creating a challenge in balancing supply and demand.

B) Consumer characteristics and product attributes

Explanation: Price customization is achieved by considering consumer characteristics and product attributes, allowing businesses to tailor prices to better match the perceived value for different segments of consumers.

B) It is easy to estimate and measure

Explanation: Cost-plus pricing is advantageous because it simplifies the pricing process, making it easy to estimate costs and justify prices to stakeholders.

B) TN / (TN + FP)

Explanation: Specificity is calculated as the ratio of true negatives (TN) to the sum of true negatives and false positives (TN + FP), measuring the test's ability to correctly identify negative cases.

B) Independent variables

Explanation: The modeling will utilize available independent variables to analyze how they influence house prices, which is crucial for accurate predictions.

B) Standardization or Inter Quartile Range

Explanation: Outliers can be treated using methods like standardization or the Inter Quartile Range, which helps in managing extreme values effectively.

C) It shrinks them towards zero

Explanation: LASSO regression shrinks the coefficients of less important predictors towards zero, effectively reducing their influence on the model and allowing for a more parsimonious model.

B) It reduces the number of predictors by shrinking some coefficients to zero

Explanation: The L1 regularization term in LASSO regression helps to reduce the number of predictors by shrinking some coefficients to zero, effectively performing variable selection and simplifying the model.

B) You will not be asked to derive

Explanation: The announcement specifies that students will not be asked to derive in the quizzes, indicating a focus on understanding rather than derivation.

B) Through surveys

Explanation: Surveys are one method to gauge customers' willingness to pay, although the text cautions that businesses should focus on actual purchasing behavior rather than just survey responses.

B) The threshold at which Youden’s index is maximum

Explanation: The optimal threshold for classification is defined as the point at which Youden’s index reaches its maximum value, indicating the best balance between sensitivity and specificity.

C) More money comes into the business than goes out

Explanation: A positive Customer Lifetime Value (LTV) indicates that over the course of the relationship, the customer generates more revenue for the business than the costs incurred, making them a valuable asset.

B) Customer interaction history

Explanation: To build an effective classification model, data on customer interaction history is essential, as it provides insights into past behaviors and responses that can inform predictions.

B) Higher return on investment

Explanation: Customizing prices based on the model can lead to higher returns, as it allows management to set prices that align with market conditions and customer demand.

C) 2^p

Explanation: With 'p' features, the number of possible subsets of predictors is 2^p, which indicates the exponential growth of combinations as the number of features increases.

C) Not knowing consumer purchase behavior when Groupons are not used

Explanation: Groupon's lack of understanding of consumer purchase behavior when Groupons are not utilized was highlighted as a critical issue in their data management.

B) Maximum Likelihood Estimation

Explanation: The content specifies that the approach for estimating regression coefficients is Maximum Likelihood Estimation, which is a common method used in statistical modeling.

B) The proportion of variability in Y explained by X

Explanation: R-squared indicates the proportion of variability in the dependent variable (Y) that can be explained by the independent variable (X), providing insight into the model's explanatory power.

B) A technique to group data into intervals

Explanation: Binning is a technique in data transformation that involves grouping data into intervals or bins, which can help in reducing noise and making patterns more apparent.

B) Domain expertise

Explanation: Domain expertise is crucial in data cleaning to accurately correct or remove errors, ensuring that the data remains relevant and valid for analysis.

C) Good fit with a large proportion explained

Explanation: An R-squared value close to 1 suggests that the regression model explains a large proportion of the variability in the dependent variable, indicating a good fit.

B) Creating additional variables based on existing data

Explanation: Constructing new variables involves creating additional variables based on existing data, which can provide new insights and improve the analysis.

B) $81,500

Explanation: The total Economic Value (EVC) for the new product is calculated as $75,000 + $19,000 - $12,500, resulting in $81,500, which reflects the overall economic assessment of the server alternatives.

B) Due to advancements in technology and data availability

Explanation: The resurgence of interest in AI is attributed to significant technological advancements and the increased availability of data, making it more applicable and beneficial for businesses today.

C) The general zoning classification of the sale

Explanation: 'MSZoning' is used to identify the general zoning classification of the sale, such as Agriculture, Commercial, or Residential Medium Density, which is crucial for understanding property regulations.

C) To standardize the range of independent variables

Explanation: Scaling is used in data transformation to standardize the range of independent variables, ensuring that each variable contributes equally to the analysis.

C) It is an art with no clear methodology

Explanation: Data preprocessing is characterized as an art rather than a science, indicating that there is no one-size-fits-all methodology and that it often develops with experience.

C) It goes to the best products, which keep changing over time

Explanation: In a hyper-competitive market, consumer demand is dynamic and shifts towards the best-performing products, necessitating companies to continuously adapt and innovate.

B) To minimize the sum of squared differences and include an L1 regularization term

Explanation: The objective of LASSO regression is to find the values of the coefficients that minimize the sum of the squared differences between predicted and actual values, while also incorporating an L1 regularization term to prevent overfitting.

C) The number of small children in the household

Explanation: The 'Kidhome' attribute specifically refers to the number of small children in the customer's household, which is a characteristic used for predictive analysis.

B) To maximize profit by aligning price with consumer value

Explanation: The main goal of price customization is to maximize profit by aligning the price of a product with the perceived value that different consumers derive from it, based on their characteristics and usage.

B) Recommendation systems

Explanation: Recommendation systems are considered a more effective approach than simple search functionalities, as they can personalize product suggestions based on user preferences and behaviors, thereby improving customer satisfaction.

A) 𝛽0, 𝛽1, 𝛽2, ..., 𝛽n

Explanation: In Multiple Linear Regression, the model parameters are represented as 𝛽0, 𝛽1, 𝛽2, ..., 𝛽n, where each 𝛽 represents a coefficient for the corresponding predictor variable.

A) The likelihood function

Explanation: L(β0, β1) denotes the likelihood function in logistic regression, which is used to find the coefficients that maximize the likelihood of observing the given data.

B) Total Sum of Squares

Explanation: TSS (Total Sum of Squares) represents the total variance in the dependent variable (Y) before regression, serving as a baseline for measuring explained variance.

C) Misleading analysis results

Explanation: Outliers can skew analysis results and lead to misleading conclusions, making it essential to identify and address them during the data cleaning process.

A) Residual Sum of Squares

Explanation: RSS (Residual Sum of Squares) indicates the variance left unexplained by the regression model, highlighting the amount of variability that the model fails to account for.

C) Categorical

Explanation: Classification in supervised learning is used when the output (y) is categorical, which can be either binary or multiclass, allowing for the categorization of data points.

C) To normalize the range of data values

Explanation: Scaling is a technique used in data transformation to normalize the range of data values, ensuring that different features contribute equally to the analysis and modeling processes.

B) The distance from random classifier performance

Explanation: Youden’s index, or Youden’s J statistic, is used to identify the threshold at which the ROC curve is farthest from the random classifier performance, indicating optimal classification.

B) Identifying relevant variables from data

Explanation: Feature extraction involves identifying and selecting the most relevant variables or features from the data that will be used in modeling, which is essential for effective analysis and prediction.

C) Phone calls

Explanation: The campaign for the gold membership is planned to be conducted through phone calls to existing customers, indicating a direct and personal approach to marketing.

A) Converting continuous data into categorical data

Explanation: Binning refers to the process of converting continuous data into categorical data by grouping values into bins or intervals, which simplifies analysis.

C) No relationship between the independent and dependent variable

Explanation: A coefficient estimate of zero indicates that there is no effect of the independent variable on the dependent variable, suggesting no relationship exists.

B) They value the product's quality

Explanation: Customers are generally willing to pay a reasonable markup because they perceive value in the product, which can lead to healthy profit margins for investors.

B) Residual Sum of Squares

Explanation: In the context of LASSO regression, 'RSS' refers to the Residual Sum of Squares, which measures the discrepancy between the predicted values and the actual values, and is minimized along with the L1 regularization term.

B) The time since the last purchase

Explanation: 'Recency' typically refers to the time elapsed since the customer's last purchase, which is an important factor in customer behavior analysis.

B) What customers do based on past purchase behaviors

Explanation: The text advises businesses to focus on actual consumer behaviors rather than what customers claim they want, as past purchase behaviors reveal true preferences.

C) Getting its pricing right

Explanation: The text emphasizes that the quickest and most effective method for a company to enhance profits is by ensuring that its pricing strategy is optimal, highlighting the importance of pricing in profitability.

A) p(xi)

Explanation: In the context of the content, p(xi) represents the sigmoid function, which is used to map predicted values to probabilities in logistic regression.

C) The linear feet of street connected to the property

Explanation: 'LotFrontage' measures the linear feet of street connected to the property, which is important for assessing accessibility and potential value.

C) An increase in the dependent variable with an increase in the independent variable

Explanation: A positive coefficient in regression analysis suggests that as the independent variable increases, the dependent variable also tends to increase, indicating a direct relationship.

A) Combining multiple data points into a single summary statistic

Explanation: Aggregation involves summarizing multiple data points into a single statistic, such as calculating the mean or sum, which helps in simplifying data analysis and interpretation.

B) $19,000

Explanation: The value differential is calculated as (20% * $100,000 - 1% * $100,000), resulting in $19,000, which is a key component in determining the Economic Value.

B) To define the analysis goals

Explanation: Formulating a business question is critical as it helps define the goals of the analysis, guiding the entire process of data mining and model development.

E) High accuracy

Explanation: High accuracy is not considered an error in data cleaning; rather, it is a desirable outcome. The other options represent various types of errors that can occur in datasets.

A) False Negative

Explanation: FN stands for False Negative, indicating the cases where the test incorrectly identifies a negative condition when it is actually positive.

B) LASSO Regression

Explanation: LASSO (Least Absolute Shrinkage and Selection Operator) Regression is an alternative method that penalizes the coefficient values, helping to regularize the model complexity and perform feature selection simultaneously.

C) Values that are absent from the dataset

Explanation: Missing values refer to data points that are absent from the dataset, which can affect analysis and require specific handling during data cleaning.

B) To analyze and extract useful information from large datasets

Explanation: Data mining tasks are essential for analyzing and extracting useful information from large datasets, which can help businesses make informed decisions and improve strategies.

B) To avoid creating an alternate market with different prices

Explanation: If products are tradeable across groups, it can lead to discrepancies in pricing, resulting in an alternate market where products are sold at different prices, undermining the pricing strategy.

C) The regression did not explain much variability

Explanation: An R-squared value close to 0 indicates that the regression model has not explained much of the variability in the dependent variable, suggesting a poor fit.

C) The sum of the absolute values of the coefficients

Explanation: The L1 regularization term in LASSO is defined as λ * (|β₁| + |β₂| + ... + |βₖ|), which sums the absolute values of the coefficients, contributing to the penalty applied to the model.

B) Conceptual understanding of business problems and ML techniques

Explanation: The quizzes are designed to test students' conceptual understanding of business problems and machine learning techniques, emphasizing the application of knowledge.

B) The maximum price a customer is ready to pay

Explanation: Willingness to pay refers to the maximum price that a customer is willing to pay for a product or service, which is crucial for effective pricing strategies.

A) True Positive

Explanation: TP stands for True Positive, which refers to the cases where the test correctly identifies a positive condition.

B) +1 and -1

Explanation: In SVM, the class labels are represented as +1 and -1, which allows for a clear distinction between the two classes being classified.

C) Individual measurable properties or characteristics of the data

Explanation: Features refer to the individual measurable properties or characteristics of the data that are used in data mining tasks to build models and make predictions.

E) Collecting raw data

Explanation: Data mining tasks focus on analyzing and extracting insights from already collected data rather than the initial collection of raw data.

B) To group data into intervals or bins

Explanation: Binning is a technique used to group data into intervals or bins, which can help in reducing noise and making patterns in the data more apparent.

C) $100,000

Explanation: The cost of a system crash is stated as $100,000 for both the new product and the next best alternative, indicating a significant risk factor in the economic evaluation.

C) The vector perpendicular to the separator

Explanation: In SVM, 'W' is the vector that is perpendicular to the decision boundary (separator), which plays a key role in defining the margin between classes.

B) $15

Explanation: The operating cost per hour for the new product is given as $15, which is an important factor in calculating the overall costs associated with the server alternatives.

C) To flip houses for profit

Explanation: The main goal of Surprise Housing's business model is to purchase houses at lower prices and then sell them at higher prices, thereby generating profit through flipping.

C) It controls the amount of shrinkage applied to the coefficients

Explanation: The penalty term (λ) in LASSO regulates how much the coefficients are shrunk towards zero, thus influencing the feature selection process and model complexity.

C) To categorize data into predefined classes

Explanation: The primary purpose of classification is to assign data points to predefined categories or classes based on their features, enabling decision-making and predictions.

B) Learning linear decision surfaces

Explanation: Before 1980, almost all learning methods focused on learning linear decision surfaces, which were known for their strong theoretical properties.

C) Build a classifier to identify valued customers

Explanation: The text suggests that developing a classifier to identify which valued customers are likely to use the coupons would be a more effective strategy for Groupon.

B) The present condition of the material on the exterior

Explanation: 'ExterCond' evaluates the present condition of the material on the exterior of the house, which is essential for assessing the property's maintenance and potential repair needs.

D) 11.1%

Explanation: A 1% improvement in price, while keeping volume constant, leads to an 11.1% increase in operating profit, indicating that pricing adjustments can have a substantial effect on profitability.

B) To shrink coefficient values to zero for feature selection

Explanation: LASSO is designed to shrink the coefficients of less important features to zero, effectively performing feature selection and simplifying the model.

C) It tests the null hypothesis that the coefficient is equal to zero

Explanation: The p-value helps determine whether the coefficient is statistically significant, indicating whether to reject the null hypothesis that the coefficient is zero, which would imply no effect.

B) Strong theoretical properties for non-linear functions

Explanation: The 1990s introduced efficient learning algorithms for non-linear functions that were based on computational learning theory and had strong theoretical properties.

A) Lifetime Value

Explanation: LTV refers to Lifetime Value, which is a crucial metric used to estimate the total revenue a business can expect from a customer throughout their relationship with the company.

B) Ordinary Least Squares

Explanation: The Ordinary Least Squares (OLS) method is commonly used to obtain coefficient estimates in Multiple Linear Regression, as it minimizes the sum of the squared differences between observed and predicted values.

B) Logistic Regression

Explanation: Logistic Regression is a widely used statistical method for binary classification tasks, predicting the probability of a binary outcome based on one or more predictor variables.

B) (R - C) * (1 + δ) / (1 + δ - r)

Explanation: The formula for LTV in an infinite horizon, assuming constant revenue and cost, is given by (R - C) * (1 + δ) / (1 + δ - r), which helps in estimating the long-term value of a customer.

C) It is often inaccurate

Explanation: A significant challenge with real-world data is its frequent inaccuracy, which necessitates preprocessing to ensure it is suitable for analysis.

B) It uses observable characteristics that correlate with the Economic Value to Customer (EVC)

Explanation: Consumer-based pricing involves using observable characteristics, such as demographics and gender, that correlate with the EVC to effectively segment the market and set prices accordingly.

B) Whether the consumer stays with the company until time t

Explanation: E(V_t) is contingent upon whether the consumer remains with the company until time t, which affects the expected value derived from that customer.

B) Ellipses and diamonds

Explanation: The solution space in LASSO regression is described using geometric shapes, specifically ellipses for the residual sum of squares and diamonds for the L1 constraint, where the optimal solution is found at the point of contact.

B) L1 regularization

Explanation: LASSO is also known as L1 regularization, which uses the absolute values of the coefficients to impose a penalty, aiding in both feature selection and model complexity reduction.

C) F(t)

Explanation: The cumulative distribution function for the random variable T, which represents customer attrition, is denoted as F(t), indicating the probability that a customer will leave by time t.

C) They suffer from local minima

Explanation: Both the neural networks developed in the 1980s and the deep neural networks from the 2010s share the common issue of suffering from local minima, which can hinder their learning processes.

C) Linear kernel

Explanation: SVM with a linear kernel uses a linear function to separate data points, making it suitable for linearly separable data in classification tasks.

B) When there are many predictors, but only a few are significant

Explanation: LASSO is particularly useful in scenarios where there are many predictors, but only a few are significant, as it helps in selecting the most relevant variables while reducing the risk of overfitting.

B) The intercept as a separate parameter

Explanation: In SVM notation, 'b' is defined as the intercept, which is a crucial parameter for determining the position of the decision boundary.

C) It simplifies the dataset for better analysis

Explanation: Reducing categories using hierarchy or intervals simplifies the dataset, making it easier to analyze and interpret the data effectively.

C) To derive actionable insights

Explanation: The ultimate goal of the model analysis, exploration, and prediction phase is to derive actionable insights that can inform business decisions and strategies.

B) To enhance the predictive power of models

Explanation: Constructing new variables aims to create additional features that can enhance the predictive power of models, allowing for better insights and more accurate predictions.

C) Price above the cost of goods sold

Explanation: Cost-oriented pricing involves setting prices above the cost of goods sold, ensuring that the company covers its costs while aiming for profitability.

B) To ensure all customers feel valued

Explanation: Maintaining fairness for low-segment users is important to ensure that all customers feel valued and included in the product offering, which can enhance brand loyalty and customer satisfaction.

C) It is crucial for modeling house prices

Explanation: Collecting data on house sales is significant for Surprise Housing as it is essential for accurately modeling house prices, which is a key component of their investment strategy.

B) They represent the average effect of the independent variable on the dependent variable, holding other variables constant

Explanation: In multiple regression, coefficients indicate the average change in the dependent variable for a one-unit change in the independent variable, assuming all other variables are held constant.

B) Development of decision trees and neural networks

Explanation: The 1980s saw the introduction of decision trees and neural networks, which allowed for the efficient learning of non-linear decision surfaces, although they had little theoretical basis and suffered from local minima.

A) Attracting new customers

Explanation: Groupon's marketing strategy was heavily focused on attracting new customers through discount offers, which ultimately led to challenges in customer retention and profitability.

B) What strategy provides the best outcome given a situation?

Explanation: This question highlights the strategic aspect of intelligent systems, focusing on optimizing outcomes based on data-driven insights.

B) To find the best-fitting line

Explanation: The purpose of estimating coefficients in Multiple Linear Regression is to find the best-fitting line that minimizes the difference between the observed values and the predicted values, thereby accurately modeling the relationship between the dependent and independent variables.

B) Finding attributes that correlate with Economic Value to Customer (EVC)

Explanation: Identifying attributes that correlate with EVC is crucial for effective pricing strategies, as it helps in understanding how customers perceive the value of different product features.

B) The probability that a customer attrits after time t

Explanation: S(t) is defined as the probability that the customer attrits after time t, calculated as S(t) = P(T > t) = 1 – P(T ≤ t) = 1 − F(t).

B) It reduces the risk of overfitting

Explanation: Regularization techniques like LASSO help in reducing the complexity of the model, which in turn lowers the risk of overfitting by preventing the model from fitting noise in the training data.

B) Matching demand with quality suppliers

Explanation: Amazon's business model emphasizes the importance of matching demand with quality suppliers, ensuring that customers can find the products they want while also supporting suppliers who may have lower visibility.

B) Close to the lift value of 2

Explanation: The queries indicate a need for probability values that are close to the lift value of 2, suggesting a specific requirement for analysis.

B) The proportion of actual negatives incorrectly identified as positives

Explanation: The false positive rate is defined as 1 - specificity, representing the proportion of actual negative cases that are incorrectly classified as positive, which is crucial for evaluating model performance.

B) The time until customer attrition

Explanation: In the context of Customer Lifetime Value, T is defined as the random variable representing the time until a customer leaves or attrits from the company.

C) LTV = σ E(V_t) / (1 + δ)^t

Explanation: The formula for calculating Customer Lifetime Value (LTV) incorporates the expected value of future profits discounted by the rate δ, emphasizing the time value of money in customer profitability analysis.

C) They allow extremely efficient learning of non-linear decision surfaces

Explanation: Deep neural networks, which emerged in the 2010s, enable extremely efficient learning of non-linear decision surfaces, although they still lack a strong theoretical basis and suffer from local minima.

B) To identify group members clearly

Explanation: The primary goal of using observable characteristics is to clearly identify group members, allowing for more tailored pricing strategies that reflect the EVC of different consumer segments.

B) The likelihood of customer attrition at a specific time

Explanation: The probability density function f(t) represents the likelihood of customer attrition occurring at a specific time t, providing insights into customer behavior over time.

B) Including a subscription price and authentication functionality in software products

Explanation: This example illustrates how product attributes can be designed to reflect customer value, allowing for differentiation in pricing based on the perceived value of the product.

B) Classify as 1 if P(ŷi) > 0.5

Explanation: The content states that by default, ŷi is classified as 1 if P(ŷi) is greater than 0.5, which is a standard threshold in logistic regression for binary classification.

D) Reinforcement Learning

Explanation: Reinforcement Learning is characterized by learning through trial and error, where an agent learns to make decisions by receiving rewards or penalties based on its actions.

B) To ensure data accuracy and reliability

Explanation: The primary purpose of data cleaning is to ensure that the data is accurate and reliable, which is crucial for effective analysis and decision-making.

B) The optimal values of the coefficients

Explanation: The contact point between the ellipse and diamond in LASSO regression represents the optimal values of the coefficients, where the model achieves the best trade-off between fitting the data and maintaining simplicity through regularization.

D) Revenue minus cost

Explanation: In the LTV formula, (R - C) represents the difference between revenue (R) and cost (C), indicating the profit contribution from a customer over time.

C) S(t) = 1 - F(t)

Explanation: S(t) is calculated as S(t) = P(T > t) = 1 – P(T ≤ t) = 1 − F(t), indicating the probability that a customer will attrit after time t.

A) Y = 𝛽0 + 𝛽1X1 + 𝛽2X2 + ... + 𝛽nXn

Explanation: The prediction equation in Multiple Linear Regression is structured as Y = 𝛽0 + 𝛽1X1 + 𝛽2X2 + ... + 𝛽nXn, where Y is the predicted value based on the linear combination of the predictors.

E) True Neutral (TN)

Explanation: True Neutral is not a recognized component of the confusion matrix. The standard components include True Positive (TP), True Negative (TN), False Positive (FP), and False Negative (FN).

B) No, they are not group work

Explanation: The announcement explicitly states that Programming Assignments are not group work, emphasizing individual effort in completing assignments.

C) Herbert A. Simon

Explanation: Herbert A. Simon made this statement in 1965, emphasizing the potential of machines and AI to perform tasks traditionally done by humans, which is a foundational concept in the field of AI.

B) To evaluate the performance of a diagnostic test

Explanation: Youden’s index is a statistic used to assess the effectiveness of a diagnostic test, combining sensitivity and specificity to provide a single measure of performance.

B) Price ≤ Value - Value of next best alternative + Price of next best alternative

Explanation: The formula indicates that the price of the product should be less than or equal to the value of the product minus the value of the next best alternative, plus the price of that alternative, ensuring it reflects the economic value.