Numpy

리스트 끼리 곱

import numpy as np

x_input=np.array([1,2,3,4,5,6,7,8,9,10],dtype = np.float32)

labels = np.array([3,5,7,9,11,13,15,17,19,21],dtype = np.float32) #

W,B = np.random.normal(size=1,), np.random.normal(size=1,)

def Hypothesis(x):

    return W*x +B

def Cost() :

    return np.mean((Hypothesis(x_input)-labels)**2)

def Gradient(x,y):

    return np.mean(x*(x+W+(B-y))), np.mean((W*x -y +B))

epochs = 5000

learning_rate = 0.005

for cnt in range(0, epochs+1) :

    if cnt % (epochs//20) ==0:

        print("[{}]cost={},W={}, B={}".format(cnt,Cost(),W,B))

    grad_W, grad_B = Gradient(x_input,labels)

    W -= learning_rate * grad_W

    B -= learning_rate * grad_B

 

함수 보면 가능.