hillclimber.java

数据挖掘聚类算法：bayes源代码,使用JAVA语言实现

	 * up to the maximum nr of parents).
	 * @throws Exception if something goes wrong
	 */
	Operation findBestArcToAdd(BayesNet bayesNet, Instances instances, Operation oBestOperation) throws Exception {
		int nNrOfAtts = instances.numAttributes();
		// find best arc to add
		for (int iAttributeHead = 0; iAttributeHead < nNrOfAtts; iAttributeHead++) {
			if (bayesNet.getParentSet(iAttributeHead).getNrOfParents() < m_nMaxNrOfParents) {
				for (int iAttributeTail = 0; iAttributeTail < nNrOfAtts; iAttributeTail++) {
					if (addArcMakesSense(bayesNet, instances, iAttributeHead, iAttributeTail)) {
						Operation oOperation = new Operation(iAttributeTail, iAttributeHead, Operation.OPERATION_ADD);
						double fScore = calcScoreWithExtraParent(oOperation.m_nHead, oOperation.m_nTail);
						if (fScore > oBestOperation.m_fScore) {
							if (isNotTabu(oOperation)) {
								oBestOperation = oOperation;
								oBestOperation.m_fScore = fScore;
							}
						}
					}
				}
			}
		}
		return oBestOperation;
	} // findBestArcToAdd

	/** 
	 * find best (or least bad) arc deletion operation
	 * 
	 * @param bayesNet Bayes network to delete arc from
	 * @param instances data set
	 * @param oBestOperation
	 * @return Operation containing best arc to delete, or null if no deletion can be made 
	 * (happens when there is no arc in the network yet).
	 * @throws Exception of something goes wrong
	 */
	Operation findBestArcToDelete(BayesNet bayesNet, Instances instances, Operation oBestOperation) throws Exception {
		int nNrOfAtts = instances.numAttributes();
		// find best arc to delete
		for (int iNode = 0; iNode < nNrOfAtts; iNode++) {
			ParentSet parentSet = bayesNet.getParentSet(iNode);
			for (int iParent = 0; iParent < parentSet.getNrOfParents(); iParent++) {
				Operation oOperation = new Operation(parentSet.getParent(iParent), iNode, Operation.OPERATION_DEL);
				double fScore = calcScoreWithMissingParent(oOperation.m_nHead, oOperation.m_nTail);
				if (fScore > oBestOperation.m_fScore) {
					if (isNotTabu(oOperation)) {
						oBestOperation = oOperation;
						oBestOperation.m_fScore = fScore;
					}
				}
			}
		}
		return oBestOperation;
	} // findBestArcToDelete

	/** 
	 * find best (or least bad) arc reversal operation
	 * 
	 * @param bayesNet Bayes network to reverse arc in
	 * @param instances data set
	 * @param oBestOperation
	 * @return Operation containing best arc to reverse, or null if no reversal is allowed
	 * (happens if there is no arc in the network yet, or when any such reversal introduces
	 * a cycle).
	 * @throws Exception if something goes wrong
	 */
	Operation findBestArcToReverse(BayesNet bayesNet, Instances instances, Operation oBestOperation) throws Exception {
		int nNrOfAtts = instances.numAttributes();
		// find best arc to reverse
		for (int iNode = 0; iNode < nNrOfAtts; iNode++) {
			ParentSet parentSet = bayesNet.getParentSet(iNode);
			for (int iParent = 0; iParent < parentSet.getNrOfParents(); iParent++) {
				int iTail = parentSet.getParent(iParent);
				// is reversal allowed?
				if (reverseArcMakesSense(bayesNet, instances, iNode, iTail) && 
				    bayesNet.getParentSet(iTail).getNrOfParents() < m_nMaxNrOfParents) {
					// go check if reversal results in the best step forward
					Operation oOperation = new Operation(parentSet.getParent(iParent), iNode, Operation.OPERATION_REVERSE);
					double fScore = calcScoreWithReversedParent(oOperation.m_nHead, oOperation.m_nTail);
					if (fScore > oBestOperation.m_fScore) {
						if (isNotTabu(oOperation)) {
							oBestOperation = oOperation;
							oBestOperation.m_fScore = fScore;
						}
					}
				}
			}
		}
		return oBestOperation;
	} // findBestArcToReverse
	

	/**
	 * Sets the max number of parents
	 *
	 * @param nMaxNrOfParents the max number of parents
	 */
	public void setMaxNrOfParents(int nMaxNrOfParents) {
	  m_nMaxNrOfParents = nMaxNrOfParents;
	} 

	/**
	 * Gets the max number of parents.
	 *
	 * @return the max number of parents
	 */
	public int getMaxNrOfParents() {
	  return m_nMaxNrOfParents;
	} 

	/**
	 * Returns an enumeration describing the available options.
	 *
	 * @return an enumeration of all the available options.
	 */
	public Enumeration listOptions() {
		Vector newVector = new Vector(2);

		newVector.addElement(new Option("\tMaximum number of parents", "P", 1, "-P <nr of parents>"));
		newVector.addElement(new Option("\tUse arc reversal operation.\n\t(default false)", "R", 0, "-R"));
		newVector.addElement(new Option("\tInitial structure is empty (instead of Naive Bayes)", "N", 0, "-N"));

		Enumeration enu = super.listOptions();
		while (enu.hasMoreElements()) {
			newVector.addElement(enu.nextElement());
		}
		return newVector.elements();
	} // listOptions

	/**
	 * Parses a given list of options. <p/>
	 *
	 <!-- options-start -->
	 * Valid options are: <p/>
	 * 
	 * <pre> -P &lt;nr of parents&gt;
	 *  Maximum number of parents</pre>
	 * 
	 * <pre> -R
	 *  Use arc reversal operation.
	 *  (default false)</pre>
	 * 
	 * <pre> -N
	 *  Initial structure is empty (instead of Naive Bayes)</pre>
	 * 
	 * <pre> -mbc
	 *  Applies a Markov Blanket correction to the network structure, 
	 *  after a network structure is learned. This ensures that all 
	 *  nodes in the network are part of the Markov blanket of the 
	 *  classifier node.</pre>
	 * 
	 * <pre> -S [LOO-CV|k-Fold-CV|Cumulative-CV]
	 *  Score type (LOO-CV,k-Fold-CV,Cumulative-CV)</pre>
	 * 
	 * <pre> -Q
	 *  Use probabilistic or 0/1 scoring.
	 *  (default probabilistic scoring)</pre>
	 * 
	 <!-- options-end -->
	 *
	 * @param options the list of options as an array of strings
	 * @throws Exception if an option is not supported
	 */
	public void setOptions(String[] options) throws Exception {
		setUseArcReversal(Utils.getFlag('R', options));

		setInitAsNaiveBayes (!(Utils.getFlag('N', options)));
		
		String sMaxNrOfParents = Utils.getOption('P', options);
		if (sMaxNrOfParents.length() != 0) {
		  setMaxNrOfParents(Integer.parseInt(sMaxNrOfParents));
		} else {
		  setMaxNrOfParents(100000);
		}
		
		super.setOptions(options);
	} // setOptions

	/**
	 * Gets the current settings of the search algorithm.
	 *
	 * @return an array of strings suitable for passing to setOptions
	 */
	public String[] getOptions() {
		String[] superOptions = super.getOptions();
		String[] options = new String[7 + superOptions.length];
		int current = 0;
		if (getUseArcReversal()) {
		  options[current++] = "-R";
		}
		
		if (!getInitAsNaiveBayes()) {
		  options[current++] = "-N";
		} 

		options[current++] = "-P";
		options[current++] = "" + m_nMaxNrOfParents;

		// insert options from parent class
		for (int iOption = 0; iOption < superOptions.length; iOption++) {
			options[current++] = superOptions[iOption];
		}

		// Fill up rest with empty strings, not nulls!
		while (current < options.length) {
			options[current++] = "";
		}
		return options;
	} // getOptions

	/**
	 * Sets whether to init as naive bayes
	 *
	 * @param bInitAsNaiveBayes whether to init as naive bayes
	 */
	public void setInitAsNaiveBayes(boolean bInitAsNaiveBayes) {
	  m_bInitAsNaiveBayes = bInitAsNaiveBayes;
	} 

	/**
	 * Gets whether to init as naive bayes
	 *
	 * @return whether to init as naive bayes
	 */
	public boolean getInitAsNaiveBayes() {
	  return m_bInitAsNaiveBayes;
	} 

	/** get use the arc reversal operation
	 * @return whether the arc reversal operation should be used
	 */
	public boolean getUseArcReversal() {
		return m_bUseArcReversal;
	} // getUseArcReversal

	/** set use the arc reversal operation
	 * @param bUseArcReversal whether the arc reversal operation should be used
	 */
	public void setUseArcReversal(boolean bUseArcReversal) {
		m_bUseArcReversal = bUseArcReversal;
	} // setUseArcReversal

	/**
	 * This will return a string describing the search algorithm.
	 * @return The string.
	 */
	public String globalInfo() {
	  return "This Bayes Network learning algorithm uses a hill climbing algorithm " +
	  "adding, deleting and reversing arcs. The search is not restricted by an order " +
	  "on the variables (unlike K2). The difference with B and B2 is that this hill " +	  
          "climber also considers arrows part of the naive Bayes structure for deletion.";
	} // globalInfo

	/**
	 * @return a string to describe the Use Arc Reversal option.
	 */
	public String useArcReversalTipText() {
	  return "When set to true, the arc reversal operation is used in the search.";
	} // useArcReversalTipText

	/**
	 * Returns the revision string.
	 * 
	 * @return		the revision
	 */
	public String getRevision() {
	  return RevisionUtils.extract("$Revision: 1.9 $");
	}
} // HillClimber