以前在统计学的学习中,有回归分析,如果只包括一个自变量和一个因变量,且二者的关系可用一条直线近似表示,这种回归分析称为一元线性回归分析,最小二乘法可根据给定的数据拟合出一条近似的直线。
package cn.zhf.test;
/**
* 最小二乘法 线性回归 y = a*x + b
*
* b = sum( y ) / n - a * sum( x ) / n
*
* a = ( n * sum( xy ) - sum( x ) * sum( y ) ) / ( n * sum( x^2 ) - sum(x) ^ 2 )
*
*/
public class LinearRegression {
public static void main(String[] args) {
int n = 0;
double[] x = { 1, 2, 3 };
double[] y = { 3, 6, 9 };
// 计算总和
double sumx = 0.0, sumy = 0.0, sumx2 = 0.0;
while (n < x.length) {
sumx += x[n];
sumx2 += x[n] * x[n];
sumy += y[n];
n++;
}
// 求平均数
double xbar = sumx / n;
double ybar = sumy / n;
// 计算系数
double xxbar = 0.0, yybar = 0.0, xybar = 0.0;
for (int i = 0; i < n; i++) {
xxbar += (x[i] - xbar) * (x[i] - xbar);
yybar += (y[i] - ybar) * (y[i] - ybar);
xybar += (x[i] - xbar) * (y[i] - ybar);
}
double beta1 = xybar / xxbar;
double beta0 = ybar - beta1 * xbar;
System.out.println("y = " + beta1 + " * x + " + beta0);
}
}