机器学习_梯度下降法一维函数

一维函数使用梯度下降法的例子

import numpy as np
import matplotlib as mpl
import matplotlib.pyplot as plt

def f1(x):
    return 0.5 * (x - 0.25) ** 2

def h1(x):
    return 0.5 * 2 *(x - 0.25)

if __name__ == '__main__':
    mpl.rcParams['font.family'] = 'SimHei'
    mpl.rcParams['axes.unicode_minus'] = False
    GD_X = []
    GD_Y = []
    X = 4
    alpha = 1.5
    f_cur = f1(X)
    f_last = f_cur
    f_change = 1
    GD_X.append(X)
    GD_Y.append(f_cur)
    #
    iter_num = 0
    while f_change > 1e-10 and iter_num < 100:
        iter_num += 1
        X = X - alpha * h1(X)
        f_cur = f1(X)
        f_change = np.abs(f_last - f_cur)
        f_last = f_cur
        GD_X.append(X)
        GD_Y.append(f_cur)
    XX = np.arange(-4, 4.5, 0.05)
    # print(list(map(lambda x: f1(x), X)))
    Y = np.array(list(map(lambda x: f1(x), XX)))
    plt.figure(facecolor='w')
    plt.plot(XX, Y, 'r-', linewidth=2)
    plt.plot(GD_X, GD_Y, 'bo--', linewidth=2)
    plt.title('$x^2$')
    plt.show()