增加隐式反馈的svd 推荐

基于svd++的java代码实现，实现了评分矩阵分解的参数计算，使用随机梯度下降，计算参数。

参考：

https://www.cnblogs.com/Xnice/p/4522671.html

https://blog.csdn.net/zhongkejingwang/article/details/43083603

推荐系统-技术、评估及高效算法

import java.io.BufferedReader;
import java.io.File;
import java.io.FileReader;
import java.io.IOException;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;

public class SVDPlus {

	//定义参数
	private double mean;
	private double[] bu;
	private double[] bi;
	private double[][] pu;
	private double[][] qi;
	private double[][] y;
	
	private int userNum;
	private Map<String,Integer> userMap = new HashMap<String, Integer>();
	private Map<Integer,String> userIndexMap = new HashMap<Integer,String> ();
	private int itemNum;
	private Map<String,Integer> itemMap = new HashMap<String, Integer>();
	private Map<Integer,String> itemIndexMap = new HashMap<Integer,String> ();
	//定义超参数
	private double learningrate;
	private double lamda1;
	private double lamda2;
	private int iter;
	private int dim;
	
	public SVDPlus(double lr,double coff1,double coff2, int it,int d){
		learningrate = lr;
		lamda1 = coff1;
		lamda2 = coff2;
		iter = it;
		dim = d;
	}
	//初始化参数
	publi