[機器學習]用Java實現梯度下降

這是在coursea的解釋 多元的梯度下降 運算過程

實際也就是求偏導數

本測試用例爲2元

但適用於多元的數據

數據如下

X

1,2,3

Y

1,2,3

代碼如下

package hello;

import java.io.BufferedReader;
import java.io.File;
import java.io.FileReader;
import java.io.IOException;
import java.util.List;
import java.util.Vector;

public class GradientDescent {
	public static Double[] getTheta(List<Double[]> X, Double[] y) {
		//初始化長度
		int m = y.length;
		//初始化theta
		Double[] theta = new Double[X.size()];
		double a = 0.001;
		for (int i = 0; i < theta.length; i++) {
			theta[i] = 0.0;
		}
		//迭代150000次
		for (int i = 0; i < 150000; i++) {
			//初始化temp,做替換用
			Double[] temp = new Double[theta.length];
			for (int j = 0; j < temp.length; j++) {
				temp[j] = 0.0;
			}
			for (int j = 0; j < m; j++) {
				Double sum = 0.0;
				for (int k = 0; k < theta.length; k++) {
					//在二元圖形中,這裏相當於k*x+b*1,三元相當於a*x+b*y+c*1,以此類推
					sum += theta[k] * X.get(k)[j];
				}
				sum = sum - y[j];
				for (int k = 0; k < theta.length; k++) {
					temp[k] += sum * X.get(k)[j];
				}
			}
			for (int j = 0; j < theta.length; j++) {
				//一起替換 同時更新
				theta[j] -= a / m * temp[j];
			}
		}
		return theta;
	}

	public static void main(String[] args) throws IOException {
		Double[] x1 = GradientDescent.read("C:/Users/ojama/Desktop/testX.txt");
		Double[] y = GradientDescent.read("C:/Users/ojama/Desktop/testY.txt");
		int m = y.length;
		Double[] x0 = new Double[m];
		for (int i = 0; i < x0.length; i++) {
			x0[i] = 1.0;
		}
		List<Double[]> X = new Vector<Double[]>();
		X.add(x0);
		X.add(x1);
		Double[] theta = GradientDescent.getTheta(X, y);
		for (int i = 0; i < theta.length; i++) {
			System.out.println(theta[i]);
		}
	}

	public static Double[] read(String fileName) throws IOException {
		File file = new File(fileName);
		FileReader fileReader = new FileReader(file);
		BufferedReader reader = new BufferedReader(fileReader);
		StringBuilder sb = new StringBuilder();
		String str = reader.readLine();
		while (str != null) {
			sb.append(str);
			str = reader.readLine();
		}
		reader.close();
		fileReader.close();
		String[] X0 = sb.toString().replace(" ", "").split(",");
		Double[] x0 = new Double[X0.length];
		for (int i = 0; i < x0.length; i++) {
			x0[i] = Double.parseDouble(X0[i]);
		}
		return x0;
	}
}

運行結果：

theta0約等於0theta1約等於1;

雖然有一定的誤差 但已經足夠精確了。