딥러닝) 신경망 학습 전체 알고리즘 (코드)

 

1. 신경망 구축에 필요한 Layers

import numpy as np
from common.functions import *


class Relu:
    def __init__(self):
        self.mask = None

    def forward(self, x):
        self.mask = (x <= 0)
        out = x.copy()
        out[self.mask] = 0

        return out

    def backward(self, dout):
        dout[self.mask] = 0
        dx = dout

        return dx


class Sigmoid:
    def __init__(self):
        self.out = None

    def forward(self, x):
        out = sigmoid(x)
        self.out = out
        return out

    def backward(self, dout):
        dx = dout * (1.0 - self.out) * self.out

        return dx


class Affine:
    def __init__(self, W, b):
        self.W = W
        self.b = b

        self.x = None
        self.original_x_shape = None
        #매개변수의 미분
        self.dW = None
        self.db = None

    def forward(self, x):
        # 텐서 대응
        self.original_x_shape = x.shape
        x = x.reshape(x.shape[0], -1)
        self.x = x

        out = np.dot(self.x, self.W) + self.b

        return out

    def backward(self, dout):
        dx = np.dot(dout, self.W.T)
        self.dW = np.dot(self.x.T, dout)
        self.db = np.sum(dout, axis=0)

        dx = dx.reshape(*self.original_x_shape)  #입력 데이터 모양 변경(텐서 대응)
        return dx


class SoftmaxWithLoss:
    def __init__(self):
        self.loss = None
        self.y = None  
        self.t = None  

    def forward(self, x, t):
        self.t = t
        self.y = softmax(x)
        self.loss = cross_entropy_error(self.y, self.t)

        return self.loss

    def backward(self, dout=1):
        batch_size = self.t.shape[0]
        if self.t.size == self.y.size:  #one-hot-encoding으로 되어있는 경우
            dx = (self.y - self.t) / batch_size
        else:
            dx = self.y.copy()
            dx[np.arange(batch_size), self.t] -= 1
            dx = dx / batch_size

        return dx

 

 

 

 

 

 

 

 

2. 신경망 모델 구축하기

import sys, os
sys.path.append(os.pardir)
import numpy as np
from common.layers import *
from common.gradient import numerical_gradient
from collections import OrderedDict

class TwoLayerNet:
    def __init__(self, input_size, hidden_size, output_size, weight_init_std=0.01): #각 layer의 노드 개수와 가중치 초기값 설정시 표준편차

        #가중치 초기화
        self.params={}
        self.params['W1'] = weight_init_std * np.random.randn(input_size, hidden_size) #평균이 0이고 표준편차가 0.01인 정규분포를 따라 랜덤 추출, 형상은 (input_size x hidden_size)
        self.params['b1'] = np.zeros(hidden_size) #zero벡터 생성
        self.params['W2'] = weight_init_std * np.random.randn(hidden_size, output_size)
        self.params['b2'] = np.zeros(output_size)

        #Layer 생성
        self.layers = OrderedDict()  #순서가 있는 딕셔너리이므로 순서가 정해짐
        self.layers['Affine1'] = Affine(self.params['W1'], self.params['b1'])
        self.layers['Relu1'] = Relu()
        self.layers['Affine2'] = Affine(self.params['W2'], self.params['b2'])

        self.lastLayer = SoftmaxWithLoss()

    def predict(self, x):
        for layer in self.layers.values():
            x = layer.forward(x)
        return x

    def loss(self, x, t):
        y = self.predict(x)
        return self.lastLayer.forward(y, t)

    def accuracy(self, x, t):
        y = self.predict(x)
        y = np.argmax(y, axis=1) #argmax는 요소가 최댓값인 index들을 리스트로 나타냄
        if t.ndim != 1 :   #ndim은 차원의 수를 나타내며 one-hot-encoding이 되어있는 경우 실행
            t = np.argmax(t, axis=1)
        accuracy = np.sum(y == t) / float(x.shape[0])   #y==t결과인 T/F에서 T의 개수를 sum(맞힌 개수), shape[0] 행의 개수(정답률)
        return accuracy


    def numerical_gradient(self, x, t):
        loss_W = lambda W: self.loss(x, t)   #람다 정규식> lambda 변수: return 식

        grads = {}
        grads['W1'] = numerical_gradient(loss_W, self.params['W1'])
        grads['b1'] = numerical_gradient(loss_W, self.params['b1'])
        grads['W2'] = numerical_gradient(loss_W, self.params['W2'])
        grads['b2'] = numerical_gradient(loss_W, self.params['b2'])
        return grads

    def gradient(self, x, t):
        #순전파
        self.loss(x, t)

        #역전파
        dout = 1 #맨 마지막 층이므로 다음 층에서 흘러들어오는 값이 없으므로 1(downstream층에서 곱하기 1하면 無의 효과)
        dout = self.lastLayer.backward(dout)

        layers = list(self.layers.values())
        layers.reverse()  #리스트의 순서 바꿈
        for layer in layers:
            dout = layer.backward(dout)

        #결과 저장
        grads = {}
        grads['W1'] = self.layers['Affine'].dW
        grads['b1'] = self.layers['Affine'].db
        grads['W2'] = self.layers['Affine'].dW
        grads['b2'] = self.layers['Affine'].db
        return grads

 

 

 

 

 

 

 

3. 데이터 학습하기

import sys, os
sys.path.append(os.pardir)
import numpy as np
from dataset.mnist import load_mnist
from two_layer_net import TwoLayerNet

(x_train, t_train), (x_test, t_test) = load_mnist(normalize = True, one_hot_label = True)
network = TwoLayerNet(input_size = 784, hidden_size = 50, output_size = 10)

iters_num = 10000
train_size = x_train.shape[0]
batch_size = 100
learning_rate = 0.1

train_loss_list = []
train_acc_list = []
test_acc_list = []

iter_per_epoch = max(train_size/batch_size, 1)

for i in range(iters_num):
    batch_mask = np.random.choice(train_size, batch_size)
    x_batch = x_train[batch_mask]
    t_batch = t_train[batch_mask]

    grad = network.gradient(x_batch, t_batch)

    for key in ('W1', 'b1', 'W2', 'b2'):
        network.params[key] -= learning_rate * grad[key]

    loss = network.loss(x_batch, t_batch)
    train_loss_list.append(loss)

    if i % iter_per_epoch == 0:
        train_acc = network. accuracy(x_train, t_train)
        test_acc = network.accuracy(x_test, t_test)
        train_acc_list.append(train_acc)
        test_acc_list.append(test_acc)
        print(train_acc, test_acc)