Package edu.cmu.tetrad.search.score

Interface AdditiveLocalScorer

All Known Implementing Classes:
CamAdditivePsplineBic
public interface AdditiveLocalScorer
Interface for scoring a target node in a graphical model based on its local structure, specifically under an additive scoring framework. This interface facilitates the calculation of scores for a node given its parent nodes using customizable parameters such as penalty discount and ridge regularization.

  • Method Summary

    Modifier and Type
    Method
    Description
    double
    localScore(int yIndex, int... parentIdxs)
    Computes the local score for a target node's index given the indices of its parent nodes in a graphical model.
    double
    localScore(Node y, Collection<Node> parents)
    Computes the local score for a target node given its set of parent nodes in a graphical model.
    default AdditiveLocalScorer
    setPenaltyDiscount(double c)
    Sets the penalty discount value used in the additive scoring framework.
    default AdditiveLocalScorer
    setRidge(double r)
    Sets the ridge regularization parameter to be used in the additive scoring framework.

  • Method Details

    • localScore

      double localScore(Node y, Collection<Node> parents)
      Computes the local score for a target node given its set of parent nodes in a graphical model.
      Parameters:
      y - the target node for which the local score is being computed
      parents - the set of parent nodes for the target node
      Returns:
      the computed local score as a double

    • localScore

      double localScore(int yIndex, int... parentIdxs)
      Computes the local score for a target node's index given the indices of its parent nodes in a graphical model.
      Parameters:
      yIndex - the index of the target node for which the local score is being computed
      parentIdxs - the indices of the parent nodes for the target node
      Returns:
      the computed local score as a double

    • setPenaltyDiscount

      default AdditiveLocalScorer setPenaltyDiscount(double c)
      Sets the penalty discount value used in the additive scoring framework.
      Parameters:
      c - the penalty discount value to be applied; this value influences the regularization applied during scoring
      Returns:
      the current instance of AdditiveLocalScorer, allowing for method chaining

    • setRidge

      default AdditiveLocalScorer setRidge(double r)
      Sets the ridge regularization parameter to be used in the additive scoring framework. The ridge parameter serves as a regularization factor to prevent overfitting and stabilize the scoring process.
      Parameters:
      r - the ridge regularization value; a higher value increases regularization, reducing the influence of highly correlated variables
      Returns:
      the current instance of AdditiveLocalScorer, allowing method chaining