entropybasedsplitcrit.java

来自「一个数据挖掘系统的源码」· Java 代码 · 共 98 行

/**
 *   
 *   AgentAcademy - an open source Data Mining framework for
 *   training intelligent agents
 *
 *   Copyright (C)   2001-2003 AA Consortium.
 *
 *   This library is open source software; you can redistribute it 
 *   and/or modify it under the terms of the GNU Lesser General 
 *   Public License as published by the Free Software Foundation;   
 *   either version 2.0 of the License, or (at your option) any later 
 *   version.
 *
 *   This library is distributed in the hope that it will be useful,
 *   but WITHOUT ANY WARRANTY; without even the implied warranty of
 *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the    
 *   GNU General Public License for more details.
 *
 *   You should have received a copy of the GNU Lesser General Public
 *   License along with this library; if not, write to the Free 
 *   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, 
 *   MA  02111-1307 USA
 * 
 */

package org.agentacademy.modules.dataminer.clusterers;
// this added by diok:
import org.agentacademy.modules.dataminer.classifiers.SplitCriterion;
/**
 * <p>Title: The Data Miner prototype</p>
 * <p>Description: A prototype for the DataMiner (DM), the Agent Academy (AA) module responsible for performing data mining on the contents of the Agent Use Repository (AUR). The extracted knowledge is to be sent back to the AUR in the form of a PMML document.</p>
 * <p>Copyright: Copyright (c) 2002</p>
 * <p>Company: CERTH</p>
 * @author asymeon
 * @version 0.3
 */

public abstract class EntropyBasedSplitCrit extends SplitCriterion{

  /** The log of 2. */
  protected static double log2 = Math.log(2);

  /**
   * Help method for computing entropy.
   */
  public final double logFunc(double num) {

    // Constant hard coded for efficiency reasons
    if (num < 1e-6)
      return 0;
    else
      return num*Math.log(num)/log2;
  }

  /**
   * Computes entropy of distribution before splitting.
   */
  public final double oldEnt(Distribution bags) {

    double returnValue = 0;
    int j;

    for (j=0;j<bags.numClasses();j++)
      returnValue = returnValue+logFunc(bags.perClass(j));
    return logFunc(bags.total())-returnValue;
  }

  /**
   * Computes entropy of distribution after splitting.
   */
  public final double newEnt(Distribution bags) {

    double returnValue = 0;
    int i,j;

    for (i=0;i<bags.numBags();i++){
      for (j=0;j<bags.numClasses();j++)
	returnValue = returnValue+logFunc(bags.perClassPerBag(i,j));
      returnValue = returnValue-logFunc(bags.perBag(i));
    }
    return -returnValue;
  }

  /**
   * Computes entropy after splitting without considering the
   * class values.
   */
  public final double splitEnt(Distribution bags) {

    double returnValue = 0;
    int i;

    for (i=0;i<bags.numBags();i++)
      returnValue = returnValue+logFunc(bags.perBag(i));
    return logFunc(bags.total())-returnValue;
  }
}