This module introduces how AI systems learn from data and use that knowledge to make predictions, discover patterns, and improve performance. Students explore supervised learning with labeled examples, including the distinction between classification and regression problems, as well as unsupervised learning methods that uncover structure and relationships in unlabeled data. The module also examines latent and hidden variables, connecting these ideas to probabilistic models such as Bayes nets and Hidden Markov Models.

This module examines the central challenge of learning: building models that generalize effectively to new, unseen data. Students explore the concepts of overfitting, underfitting, and the bias-variance tradeoff, along with the processes involved in training and evaluating learning models. The module also introduces the roles of training, validation, and testing data sets in model development and examines the practical challenges that arise in AI learning systems, including data limitations, optimization difficulties, scalability, and changing environments.

This module introduces major families of AI and machine learning models, including linear models, decision trees, neural networks, and reinforcement learning. Students explore how each model family represents knowledge, learns from data or experience, and makes decisions or predictions. The module also connects these classical and modern learning approaches to contemporary AI systems such as large language models, recommendation systems, and robotics.