Bayesian Network Explainer
A Bayesian network is a probabilistic model whose directed acyclic graph represents variables and their conditional dependencies. Each node has a probability rule conditioned on its parent nodes. Together, the graph and those local rules define a joint probability distribution.
The factorized structure can make complex probability calculations manageable. When evidence is observed, inference updates beliefs about unobserved variables. For example, symptoms and test results can change the estimated probability of possible conditions, subject to the model’s assumptions and data.
The graph encodes conditional independence claims: once certain variables are known, others may provide no additional information within the model. This structure can come from expert knowledge, learned data relationships, or both. Different graphs can sometimes represent the same set of independencies.
An arrow does not automatically mean that one variable causes another. Causal interpretation requires stronger assumptions and a carefully justified causal model. Missing variables, inaccurate probabilities, or shifted conditions can make inferences unreliable. Results should include uncertainty and sensitivity to important assumptions.
A Bayesian network represents variables and conditional relationships in a directed acyclic graph.
Key terms include directed acyclic graph, node, edge, parent, conditional probability, joint distribution, evidence, inference and conditional independence.
Each node contributes a conditional probability given its parents, and multiplying those local terms yields the full joint distribution represented by the network.
Inference conditions that distribution on observed evidence to answer questions about remaining variables.
A dependency graph can summarize statistical relationships without proving what would happen under an intervention.
Causal claims need justified direction, relevant variables, and assumptions that go beyond fitting observed correlations.
Explore more "Explainers"
Discover additional explainers across politics, science, business, technology, and other fields. Each explainer breaks down a complex idea into clear, everyday language—helping you better understand how major concepts, systems, and debates shape the world around us.