Selecting a Form of ML CT-AI Practice Questions (Page 5)

Questions 21–25

1. 21expert · hard

   A smart home company wants to optimize energy consumption. The system can control heating, cooling, and lighting, and receives a reward for reducing energy usage while maintaining comfort. The system must learn a strategy over time by interacting with the home environment. Which form of ML is most appropriate?

   • AUnsupervised clustering to group similar usage patterns
   • BReinforcement learning to learn an energy optimization policy
   • CDimensionality reduction to simplify sensor data
   • DSupervised learning to predict energy usage from historical data
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   An AI system learns to play a board game by making moves and receiving a positive score for winning and a negative score for losing. Which type of machine learning does this represent?

   • AUnsupervised learning
   • BSupervised learning
   • CSemi-supervised learning
   • DReinforcement learning
   Reveal Solution Discussion Select an answer first
4. 23foundation · easy

   A company has a large dataset of customer purchase histories but no predefined customer segments. They want to identify natural groups of customers with similar buying patterns. Which ML form is most appropriate?

   • ASupervised learning (classification)
   • BUnsupervised learning (clustering)
   • CSupervised learning (regression)
   • DReinforcement learning
   Reveal Solution Discussion Select an answer first
5. 24application · medium

   A hospital has a dataset of patient records with known diagnoses (e.g., diabetic, healthy). The IT team wants to build a system that can classify new patients into these diagnostic categories. The team is debating between writing explicit rules and using ML. Which statement correctly justifies the choice of ML over traditional programming?

   • AML can automatically learn the mapping from patient features to diagnoses from the labeled data, without manually specifying every rule
   • BML replaces the need for any human oversight in medical decisions
   • CML guarantees 100% accurate predictions on new patients
   • DML does not require any data to build the model
   Reveal Solution Discussion Select an answer first
6. 25foundation · easy

   A team wants to build a model that predicts whether a loan applicant will default (yes/no) based on historical data where each applicant is labeled as 'defaulted' or 'not defaulted'. Which form of ML is most appropriate?

   • ASupervised learning (classification)
   • BUnsupervised learning (clustering)
   • CReinforcement learning
   • DSupervised learning (regression)
   Reveal Solution Discussion Select an answer first