Introduction to Machine Learning

What is the primary objective of the minimax algorithm in game-playing scenarios?

• Maximizing the player's score while minimizing the opponent's score

• Randomly selecting moves to confuse the opponent

• Minimizing computation time

Alpha-beta pruning improves the efficiency of the minimax algorithm by:

• Expanding all possible game trees

• Eliminating branches that cannot influence the final decision

• Increasing the depth of the search

In adversarial search, the term "utility value" refers to:

• The cost of executing a move

• A numerical value representing the desirability of a game state

• The number of players in the game

Which of the following is not a machine learning paradigm?

• Supervised learning

• Unsupervised learning

• Rule-based programming

Version spaces in machine learning are used to:

• Represent all hypotheses consistent with the training data

• Store historical versions of datasets

• Optimize neural network architectures

The PAC (Probably Approximately Correct) learning framework focuses on:

• Guaranteeing exact correctness of hypotheses

• Balancing accuracy and computational efficiency

• Ignoring training data variability

Decision trees split data based on:

• Random selection of features

• Feature thresholds that maximize information gain

• Predefined user rules

The k-means clustering algorithm groups data points by:

• Minimizing the distance between points and cluster centroids

• Maximizing variance within clusters

• Using labeled training data

Naïve Bayes classifiers assume that features are:

• Conditionally independent given the class

• Highly correlated with each other

• Irrelevant to the classification task

Which technique is used for multi-class classification?

• Linear regression

• One-vs-Rest (OvR)

• K-means clustering