Package opennlp.tools.ml.maxent.quasinewton

Class LineSearch

  • public class LineSearch
extends java.lang.Object
    Class that performs line search to find minimum

    • Nested Class Summary

      Nested Classes 
      Modifier and Type Class Description
      static class  LineSearch.LineSearchResult
      Class to store lineSearch result

    • Constructor Summary

      Constructors 
      Constructor Description
      LineSearch()  

    • Method Summary

      All Methods Static Methods Concrete Methods 
      Modifier and Type Method Description
      static void doConstrainedLineSearch​(Function function, double[] direction, LineSearch.LineSearchResult lsr, double l1Cost, double initialStepSize)
      Constrained line search (see section 3.2 in the paper "Scalable Training of L1-Regularized Log-Linear Models", Andrew et al.
      static void doLineSearch​(Function function, double[] direction, LineSearch.LineSearchResult lsr, double initialStepSize)
      Backtracking line search (see Nocedal & Wright 2006, Numerical Optimization, p.

      • Methods inherited from class java.lang.Object

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

    • Constructor Detail

      • LineSearch

        public LineSearch()

    • Method Detail

      • doLineSearch

        public static void doLineSearch​(Function function,
                                double[] direction,
                                LineSearch.LineSearchResult lsr,
                                double initialStepSize)
        Backtracking line search (see Nocedal & Wright 2006, Numerical Optimization, p. 37)

      • doConstrainedLineSearch

        public static void doConstrainedLineSearch​(Function function,
                                           double[] direction,
                                           LineSearch.LineSearchResult lsr,
                                           double l1Cost,
                                           double initialStepSize)
        Constrained line search (see section 3.2 in the paper "Scalable Training of L1-Regularized Log-Linear Models", Andrew et al. 2007)