Gradient_Descent.py

import numpy as np
import matplotlib.pyplot as plt


def draw(x1,x2):
    ln = plt.plot(x1,x2,'-')
    plt.pause(0.0001)
    ln[0].remove()


def sigmoid(score):
    return 1/(1+np.exp(-score))

def calc_error(linear_parameters,points,y):
    m = points.shape[0]
    p = sigmoid(points*linear_parameters)
    cross_entropy = -(np.log(p).T*y + np.log(1-p).T*(1-y))/m
    return cross_entropy
def gradient_descent(line_parameters,points,y,alpha):
    m = points.shape[0]
    for i in range(2000):
        p = sigmoid(points*line_parameters)
        gradient = points.T*(p-y)*(alpha/m)
        line_parameters = line_parameters-gradient
        
        w1 = line_parameters.item(0)
        w2 = line_parameters.item(1)
        b = line_parameters.item(2)
        
        x1 = np.array([points[:,0].min(),points[:,0].max()])
        x2 = -b/w2+(x1*(-w1/w2))
        draw(x1,x2)

pts=100
np.random.seed(0)
bias = np.ones(pts)
top_region = np.array([np.random.normal(10,2,pts),np.random.normal(12,2,pts),bias]).T
bottom_region = np.array([np.random.normal(5,2,pts),np.random.normal(6,2,pts),bias]).T
all_points = np.vstack((top_region,bottom_region))
linear_parameters = np.matrix(np.zeros(3)).T
'''x1 = np.array([bottom_region[:,0].min(),top_region[:,0].max()])
x2 = -b/w1+x1*(-w1/w2)'''
y = np.array([np.zeros(pts),np.ones(pts)]).reshape(pts*2,1)

plt.scatter(top_region[:,0],top_region[:,1],c='r')
plt.scatter(bottom_region[:,0],bottom_region[:,1],c='b')
gradient_descent(linear_parameters,all_points,y,0.06)
plt.show()