pecfuh.cpp

模糊K近邻分类器

//该函数实现对prototype部分的训练
/* 输入参数包括：
     prototype -- 指向prototype部分的指针
	 N -- prototype部分的结点个数
	 TrainNum -- 训练样本的个数
	 L -- 训练样本的维数
	 Train -- 训练样本集合
	 Train_out -- 训练样本的期望输出
 */
#include <iostream.h>
#include <stdio.h>
#include "FuzzyNN.h"
#include "feedfoward.h"
#include "tools.h"
#include <math.h>

void pecfuh(ProtoNode *prototype,int N,int TrainNum,int L,FuzzyNum **Train,int *Train_out)
{
	double alpha0;
	int i,j,k;
	int def=5;
	double alpha;

	cout<<"==================================="<<endl;
	cout<<"现在开始对神经网络进行第一趟的训练."<<endl;

	cout<<"请输入学习速率的大小:"<<endl;
	cin>>alpha0;

	//初始化prototype部分
	for(j=1;j<=N;j++)
	{
		for(i=1;i<=L;i++)
		{
			(prototype[j].w)[i].w1=frand(0.45,0.55);
			(prototype[j].w)[i].w2=frand(0.45,0.55);
			(prototype[j].w)[i].a=frand(0.45,0.55);
			(prototype[j].w)[i].b=frand(0.45,0.55);
		}
		for(i=1;i<=N;i++)
			if(i==j)(prototype[j].g)[i]=1;
			else (prototype[j].g)[i]=0;
	}
    //WeightOut(prototype,N,L);
	//对胜出结点的权值进行调整
	int c;
	for(k=1;k<=def;k++)
	{
		 alpha=alpha0*(1-(k-1)/(double)def);
    	 for(j=1;j<=TrainNum;j++)
		 {
     		c=feedforward(prototype,N,NULL,0,Train[j],L);
	    	if(c==Train_out[j])
			{
		    	for(i=1;i<=L;i++)
				{
			    	(prototype[c].w)[i].w1+=alpha*(Train[j][i].w1-(prototype[c].w)[i].w2);
			    	(prototype[c].w)[i].w2+=alpha*(Train[j][i].w2-(prototype[c].w)[i].w1);
			    	(prototype[c].w)[i].a+=alpha*(Train[j][i].a+(prototype[c].w)[i].b);
			    	(prototype[c].w)[i].b+=alpha*(Train[j][i].b+(prototype[c].w)[i].a);
				}
			}
	    	else
			{
		    	for(i=1;i<=L;i++)
				{
			    	(prototype[c].w)[i].w1+=-alpha*(Train[j][i].w2-(prototype[c].w)[i].w1);
			    	(prototype[c].w)[i].w2+=-alpha*(Train[j][i].w1-(prototype[c].w)[i].w2);
			    	(prototype[c].w)[i].a+=alpha*(Train[j][i].b+(prototype[c].w)[i].a);
			    	(prototype[c].w)[i].b+=alpha*(Train[j][i].a+(prototype[c].w)[i].b);
				}
			}
	      //WeightOut(prototype,N,L);
		 }
	}
	cout<<"神经网络第一趟训练完毕."<<endl;
	cout<<"==================================="<<endl;
}