Selecting a Form of ML CT-AI Practice Questions (Page 1)

Questions 1–5

1. 1application · medium

   A telecom company wants to predict customer churn (whether a customer will leave). They have historical data with labels for churned and retained customers. The team also wants to understand which factors (e.g., plan type, usage) are most influential. Which form of ML should be used?

   • ASupervised regression to predict churn probability
   • BUnsupervised clustering to group customers by churn behavior
   • CSupervised classification to predict churn and analyze feature importance
   • DReinforcement learning to optimize retention offers
   Reveal Solution Discussion Select an answer first
2. 2application · medium

   A data science team has a dataset with thousands of features (e.g., sensor readings) and wants to reduce the number of features before feeding it into a supervised classification model. The goal is to retain the most important information while simplifying the dataset. Which approach should the team use?

   • AUse a regression algorithm to predict the number of features to keep
   • BUse unsupervised dimensionality reduction (e.g., PCA) to compress the features
   • CUse reinforcement learning to learn which features are most important
   • DUse a clustering algorithm to group similar features and pick one from each group
   Reveal Solution Discussion Select an answer first
4. 3foundation · easy

   Which statement best describes the difference between machine learning and traditional programming?

   • AMachine learning and traditional programming are identical, with the only difference being the programming language used.
   • BIn traditional programming, rules and data are inputs and answers are outputs; in machine learning, data and answers are inputs and rules are outputs.
   • CTraditional programming is used only for numerical tasks, while machine learning is used for all other tasks.
   • DMachine learning requires explicit programming of every possible rule, while traditional programming can infer rules from data.
   Reveal Solution Discussion Select an answer first
5. 4expert · hard

   A retail chain wants to understand the purchasing behavior of its customers. They have transaction data with no labels. The team wants to identify distinct customer segments and also reduce the number of features (e.g., product categories) to simplify analysis. Which combination of techniques is most appropriate?

   • AUse supervised classification with manually defined segments
   • BUse reinforcement learning to optimize purchasing behavior
   • CUse dimensionality reduction first, then clustering on the reduced data
   • DUse clustering first, then dimensionality reduction on the cluster centers
   Reveal Solution Discussion Select an answer first
6. 5foundation · easy

   Which of the following is a supervised learning task?

   • AReducing the number of features in a dataset while preserving variance.
   • BTraining an agent to maximize rewards by navigating a maze.
   • CPredicting the price of a house based on its size, location, and number of bedrooms.
   • DGrouping customers into segments based on purchasing behavior without predefined categories.
   Reveal Solution Discussion Select an answer first