com.expleague.ml.cli.builders.methods.impl.RidgeRegressionLeavesObliviousTreeBuilder Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of jmll

Various ML methods implemented by myself and my students

There is a newer version: 1.4.9

package com.expleague.ml.cli.builders.methods.impl;

import com.expleague.commons.func.Factory;
import com.expleague.commons.random.FastRandom;
import com.expleague.ml.BFGrid;
import com.expleague.ml.methods.MultipleVecOptimization;
import com.expleague.ml.methods.VecOptimization;
import com.expleague.ml.methods.trees.GreedyObliviousTree;
import com.expleague.ml.methods.trees.GreedyObliviousTreeWithVecOptimizationLeaves;
import com.expleague.ml.loss.L2;
import com.expleague.ml.loss.WeightedLoss;
import com.expleague.ml.methods.linearRegressionExperiments.MultipleRidgeRegression;

/**
 * User: noxoomo
 */
public class RidgeRegressionLeavesObliviousTreeBuilder implements Factory {
  public static FastRandom defaultRandom;
  public static Factory defaultGridBuilder;
  private Factory gridBuilder = defaultGridBuilder;
  public FastRandom random = defaultRandom;
  private int depth = 6;
  private double lambda = 1.0;

  public void setLambda(final double lambda) {
    this.lambda = lambda;
  }

  public void setDepth(final int depth) {
    this.depth = depth;
  }

  public void setGridBuilder(final Factory gridBuilder) {
    this.gridBuilder = gridBuilder;
  }

  public void setRandom(final FastRandom random) {
    this.random = random;
  }

  @Override
  public VecOptimization create() {
    final GreedyObliviousTree tree = new GreedyObliviousTree<>(gridBuilder.create(),depth);
    final MultipleVecOptimization leafLearner = new MultipleRidgeRegression(lambda);
    return new GreedyObliviousTreeWithVecOptimizationLeaves(tree,leafLearner,random);
  }
}