Differentiate the Use Between Different Algorithm Types (supervised, Unsupervised, Reinforcement) CERTIFIED-AI-EXPERT Practice Questions (Page 3)

Questions 11–15

1. 11foundation · easy

   What is a key difference between reinforcement learning and supervised learning?

   • AReinforcement learning learns from a reward signal, while supervised learning learns from labeled examples.
   • BReinforcement learning uses labeled data, while supervised learning uses reward signals.
   • CBoth use labeled data, but reinforcement learning also uses an environment.
   • DReinforcement learning discovers patterns in unlabeled data, while supervised learning uses rewards.
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   A robotics company is developing a robotic arm that must learn to pick up objects of varying shapes and weights. The arm receives a positive reward when it successfully grasps an object and a negative reward when it drops it. The arm learns over time by adjusting its movements based on these rewards. Which algorithm type is being used?

   • AUnsupervised learning, because the arm discovers patterns in its own movements without external guidance.
   • BReinforcement learning, because the arm is an agent that learns through trial-and-error interactions with its environment using reward signals.
   • CSupervised learning, because the arm is trained on a dataset of successful and unsuccessful grasps.
   • DImitation learning, because the arm is mimicking a human demonstrator's movements.
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   A game development company wants to create an AI opponent that learns to play a strategy game by playing against itself millions of times. The AI receives points for winning and loses points for losing, and it adjusts its strategy to maximize its score over time. Which algorithm type is best suited for this task?

   • ASupervised learning, by training the AI on a dataset of moves made by expert human players.
   • BImitation learning, by having the AI mimic the moves of a champion player.
   • CReinforcement learning, because the AI is an agent that learns a strategy through interactions with the game environment and reward signals.
   • DUnsupervised learning, by clustering game states to find winning strategies.
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   Which scenario is an example of an unsupervised learning application?

   • ATeaching a self-driving car to stay in its lane by rewarding correct driving behavior.
   • BPredicting the likelihood of a patient having a disease based on labeled medical records.
   • CReducing the number of features in a high-dimensional dataset while preserving its variance.
   • DDetecting fraudulent transactions using a dataset where each transaction is labeled as fraudulent or not.
   Reveal Solution Discussion Select an answer first
6. 15application · medium

   A marketing team has a large database of customer purchase histories but no predefined customer segments. They want to identify distinct groups of customers with similar buying behaviors so they can tailor promotions. Which algorithm type should they use?

   • AUnsupervised learning, using clustering to discover natural groupings based on purchase patterns.
   • BReinforcement learning, by having an agent explore different promotions to maximize customer engagement.
   • CSupervised learning, by training a classifier on the purchase data to assign each customer to a known segment.
   • DDimensionality reduction, to reduce the number of features before applying any other algorithm.
   Reveal Solution Discussion Select an answer first