Module es.upm.fi.cig.multictbnc
Package es.upm.fi.cig.multictbnc.learning.parameters.bn

Class BNBayesianEstimation

java.lang.Object
es.upm.fi.cig.multictbnc.learning.parameters.bn.BNParameterLearningAlgorithm
es.upm.fi.cig.multictbnc.learning.parameters.bn.BNBayesianEstimation
All Implemented Interfaces:
ParameterLearningAlgorithm
public class BNBayesianEstimation extends BNParameterLearningAlgorithm
Implements the Bayesian estimation to estimate the parameters of a BN. It is assumed a Dirichlet prior distribution over the probabilities of each state of the variables, given the state of their parents. The imaginary counts should be defined based on the number of states of the node and its parents, so the total number of imaginary samples is not influenced by the cardinality of those nodes.

  • Constructor Details

    • BNBayesianEstimation

      public BNBayesianEstimation(double nxHP)
      Receives the hyperparameter of the Dirichlet prior distribution over the parameters (i.e. imaginary counts).
      Parameters:
      nxHP - number of times the variables are in a certain state while its parents take a certain instantiation (hyperparameter)

  • Method Details

    • getIdentifier

      public String getIdentifier()
      Description copied from interface: ParameterLearningAlgorithm
      Returns a unique identifier for the parameter learning algorithm.
      Returns:
      unique identifier for the parameter learning algorithm

    • getNameMethod

      public String getNameMethod()
      Description copied from interface: ParameterLearningAlgorithm
      Gets the name of the method to learn the parameters.
      Returns:
      name of the method to learn the parameters

    • getParametersAlgorithm

      public Map<String,String> getParametersAlgorithm()
      Description copied from interface: ParameterLearningAlgorithm
      Returns the parameters that are used by the algorithm.
      Returns:
      a Map with the parameters used by the algorithm

    • getSufficientStatisticsNode

      protected BNSufficientStatistics getSufficientStatisticsNode(DiscreteStateNode node, Dataset dataset)
      Description copied from class: BNParameterLearningAlgorithm
      Returns the sufficient statistics of a DiscreteNode for a given Dataset.
      Specified by:
      getSufficientStatisticsNode in class BNParameterLearningAlgorithm
      Parameters:
      node - a DiscreteNode
      dataset - dataset from which the sufficient statistics are extracted
      Returns:
      sufficient statistics of the provided node