Identify Supervised, Unsupervised, and Reinforcement Learning SNOWPRO-ADVANCED-DATA-SCIENTIST Practice Questions (Page 1)

Questions 1–5

1. 1expert · hard

   A gaming company is developing an AI opponent for a strategy game. The AI must learn to adapt its tactics based on the player's moves. The company has a large dataset of past game logs, but the logs do not include the 'best' move for each situation. The AI will be rewarded for winning the game. Which learning approach should the AI use?

   • ASupervised learning on the game logs
   • BSemi-supervised learning with pseudo-labels
   • CReinforcement learning with the game as the environment
   • DUnsupervised learning to cluster game states
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   What is a key characteristic of unsupervised learning?

   • AThe model is trained with input-output pairs to minimize prediction error
   • BThe model identifies patterns or groupings without using labeled outcomes
   • CThe model requires a predefined target variable for each observation
   • DThe model learns from a reward signal that indicates the quality of its actions
   Reveal Solution Discussion Select an answer first
4. 3expert · hard

   A ride-sharing company wants to develop an AI that learns to set surge pricing. The AI receives real-time demand and supply data and must decide the price multiplier. The AI is rewarded for maximizing revenue while maintaining customer satisfaction. Which learning approach is most appropriate?

   • ASupervised learning to predict demand
   • BUnsupervised learning to cluster demand patterns
   • CReinforcement learning with revenue and satisfaction as rewards
   • DSemi-supervised learning with a small labeled set
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   Which components are essential to reinforcement learning?

   • AUnlabeled data, clusters, and centroids
   • BLabeled data, features, and target variables
   • CAgent, environment, actions, and rewards
   • DInput features, output labels, and a loss function
   Reveal Solution Discussion Select an answer first
6. 5foundation · easy

   A data scientist wants to group customers into segments based on purchasing behavior, without using any predefined customer categories. Which learning paradigm should they use?

   • AReinforcement learning
   • BSupervised learning
   • CSemi-supervised learning
   • DUnsupervised learning
   Reveal Solution Discussion Select an answer first