Bayesian Networks and Decision Graphs

What you'll learn

• Construct and interpret Bayesian network structures for real-world problems
• Apply exact inference algorithms (variable elimination, junction tree) to compute probabilities
• Use approximate inference methods (MCMC, importance sampling, variational inference) for large or continuous models
• Estimate parameters and learn network structure from data, including score- and constraint-based methods
• Model hybrid and continuous variables and design dynamic Bayesian networks for temporal data
• Perform causal discovery and reason about interventions and counterfactuals
• Formulate decision problems with influence diagrams and compute optimal policies
• Implement algorithms for decision evaluation and sequential decision-making under uncertainty
• Validate models, diagnose errors, and create explainable probabilistic workflows for deployment

Prerequisites

Familiarity with basic probability, statistics, and programming (preferably Python); calculus and linear algebra recommended.

Skills you'll gain

• Probabilistic reasoning
• Bayesian inference
• Approximate inference methods
• Structure and parameter learning
• Causal inference and interventions
• Decision analysis and influence diagrams
• Dynamic and hybrid modeling
• Model validation and diagnostics
• Explainable AI for probabilistic models
• Tooling and deployment of graphical models