一眼学会python深度学习之神经网络

第一步 看一遍然后学会

问题分析
完成对手写数字数据集的读取、神经网络的推理和批处理过程，3层网络结构（第一隐藏层100个神经元，第二隐藏层50个神经元，输出层10个神经元，神经网络的参数随机生成），计算模型的准确率，并给出程序的注释。

# -*- coding: utf-8 -*-
"""
Created on FRI Sep 30 17:21:22 2022

@author: math07
"""

import numpy as np
import pickle

def load_data():
    save_file = r"D:\【源代码】深度学习入门：基于Python的理论与实现\ch03\dataset\mnist.pkl"
    with open(save_file, 'rb') as f:
        dataset = pickle.load(f) 
        
    x_train, t_train = dataset["train_img"], dataset["train_label"] 
    x_test, t_test = dataset["test_img"], dataset["test_label"] 
    return x_train, t_train,x_test, t_test


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

def softmax(x):
    x = x-np.max(x)
    exp_sum = np.sum(np.exp(x))
    
    return np.exp(x)/exp_sum 

def init_weight():
    network={}             
    network["w1"] = np.random.rand(784,50)
    network["b1"] = np.random.rand(50)
    network["w2"] = np.random.rand(50,100)
    network["b2"] = np.random.rand(100)
    network["w3"] = np.random.rand(100,10)
    network["b3"] = np.random.rand(10)
    return network
def predict(network,x):
    w1=network["w1"]
    b1=network["b1"] 
    w2=network["w2"] 
    b2=network["b2"] 
    w3=network["w3"]
    b3=network["b3"] 
    a1 = np.dot(x, w1) + b1
    z1 = sigmoid(a1)
    a2 = np.dot(z1, w2) + b2
    z2 = sigmoid(a2)
    a3 = np.dot(z2, w3) + b3
    y = softmax(a3)
    return y

b_size=100
x_train,t_train,x_test,t_test=load_data()
x=x_test[0]
accuracy =0
network =init_weight()
for i in range(0,len(x_test),b_size):
    x_b=x_test[i:i+b_size]
    y_b=predict(network,x_b)
    p=np.argmax(y_b,axis=1)
    accuracy += np.sum(p ==t_test[i:i+b_size])
print('accuracy:'+str(accuracy/10000))

第二步 沉浸在成功的喜悦

在这里插入图片描述