Class BayesianBeliefNetwork

java.lang.Object

org.episteme.core.mathematics.discrete.DirectedWeightedGraph<String,Real>

org.episteme.core.mathematics.ml.BayesianBeliefNetwork

All Implemented Interfaces:

Graph<String>, WeightedGraph<String,Real>

public class BayesianBeliefNetwork
extends DirectedWeightedGraph<String,Real>

Represents a Bayesian Belief Network (BBN). A probabilistic graphical model that represents a set of variables and their conditional dependencies via a directed acyclic graph (DAG).

Since:

1.0

Author:

Silvere Martin-Michiellot, Gemini AI (Google DeepMind)

Nested Class Summary

Nested classes/interfaces inherited from interface Graph

Graph.Edge<V>

Constructor Summary

Constructors

Constructor	Description
BayesianBeliefNetwork()

Method Summary

Modifier and Type	Method	Description
void	addDependency(String parent, String child)	Adds a dependency between two variables.
void	addVariable(String name, String... states)	Adds a random variable to the network.
Real	query(String target, String targetState, Map<String,String> evidence)	Performs exact inference using the Variable Elimination algorithm.
void	setCPT(String name, double... probabilities)	Sets the Conditional Probability Table (CPT) for a variable.

Methods inherited from class DirectedWeightedGraph

addEdge, addVertex, degree, edges, getDefaultWeight, getWeight, getWeightedEdges, getWeightedEdgesFrom, isDirected, neighbors, vertexCount, vertices

Methods inherited from class Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface Graph

bfs, dfs

Methods inherited from interface WeightedGraph

addEdge, hasEdge

Constructor Details

BayesianBeliefNetwork

public BayesianBeliefNetwork()

Method Details

addVariable

public void addVariable(String name,
 String... states)

Adds a random variable to the network.

Parameters:

name - Name of the variable.

states - Possible states of the variable.

addDependency

public void addDependency(String parent,
 String child)

Adds a dependency between two variables.

Parameters:

parent - The parent variable.

child - The child variable.

setCPT

public void setCPT(String name,
 double... probabilities)

Sets the Conditional Probability Table (CPT) for a variable.

Parameters:

name - The variable name.

probabilities - The probabilities in a specific order corresponding to the parent states combinations.

query

public Real query(String target,
 String targetState,
 Map<String,String> evidence)

Performs exact inference using the Variable Elimination algorithm.

Parameters:

target - The variable to query.

targetState - The state of interest.

evidence - A map of observed variables and their states.

Returns:

The probability P(target=targetState | evidence).