cs231n homework 两层全连接神经网络 分类CIFAR-10
cs231n homework 两层全连接神经网络 分类CIFAR-10
code
import numpy as np
import pickle
import matplotlib.pyplot as plt
def loadfile():
train_data = []
train_label = []
val_data = []
val_label = []
test_data = []
test_label = []
for i in range(1, 6):
train_file = open('F:\PycharmProjects\cs231n\\two_layer_neuralnet\cifar-10-batches-py\data_batch_'+str(i),'rb')
train_file_object = pickle.load(train_file,encoding='bytes')
for line in train_file_object[b'data']:
train_data.append(line)
for line in train_file_object[b'labels']:
train_label.append(line)
train_data = np.array(train_data)
train_label = np.array(train_label)
subtraintoval = np.random.choice(train_data.shape[0],int(0.1*train_data.shape[0]))
print(len(subtraintoval))
val_data = train_data[subtraintoval].astype("float")
val_label = train_label[subtraintoval]
train_data = np.delete(train_data,subtraintoval,0)
train_label = np.delete(train_label,subtraintoval,0)
train_data = train_data.astype("float")
test_file = open('F:\PycharmProjects\cs231n\\two_layer_neuralnet\cifar-10-batches-py\\test_batch', 'rb')
test_file_object = pickle.load(test_file, encoding='bytes')
# print(test_file_object)
for line in test_file_object[b'data']:
test_data.append(line)
for line in test_file_object[b'labels']:
test_label.append(line)
test_data = np.array(test_data).astype("float")
test_label = np.array(test_label)
print("train_data shape:" + str(train_data.shape))
print("train_label shape:" + str(train_label.shape))
print("val_data shape:" + str(val_data.shape))
print("val_label shape:" + str(val_label.shape))
print("test_data shape:" + str(test_data.shape))
print("test_label shape:" + str(test_label.shape))
return train_data,train_label,val_data,val_label,test_data,test_label
class TwoLayerNet(object):
"""
A two-layer fully-connected neural network with ReLU nonlinearity and
softmax loss that uses a modular layer design. We assume an input dimension
of D, a hidden dimension of H, and perform classification over C classes.
The architecure should be affine - relu - affine - softmax.
Note that this class does not implement gradient descent; instead, it
will interact with a separate Solver object that is responsible for running
optimization.
The learnable parameters of the model are stored in the dictionary
self.params that maps parameter names to numpy arrays.
"""
def __init__(self, input_dim=3 * 32 * 32, hidden_dim=100, num_classes=10,
weight_scale=1e-3, reg=0.0):
"""
Initialize a new network.
Inputs:
- input_dim: An integer giving the size of the input
- hidden_dim: An integer giving the size of the hidden layer
- num_classes: An integer giving the number of classes to classify
- dropout: Scalar between 0 and 1 giving dropout strength.
- weight_scale: Scalar giving the standard deviation for random
initialization of the weights.
- reg: Scalar giving L2 regularization strength.
"""
self.params = {}
self.reg = reg
############################################################################
# TODO: Initialize the weights and biases of the two-layer net. Weights #
# should be initialized from a Gaussian with standard deviation equal to #
# weight_scale, and biases should be initialized to zero. All weights and #
# biases should be stored in the dictionary self.params, with first layer #
# weights and biases using the keys 'W1' and 'b1' and second layer weights #
# and biases using the keys 'W2' and 'b2'. #
############################################################################
self.params['W1'] = weight_scale * np.random.randn(input_dim, hidden_dim)
# print(W1.shape)
self.params['b1'] = np.zeros(hidden_dim)
# print(b1.shape)
self.params['W2'] = weight_scale * np.random.randn(hidden_dim, num_classes)
self.params['b2'] = np.zeros(num_classes)
############################################################################
# END OF YOUR CODE #
############################################################################
def loss(self, X, y=None,reg =0.0):
"""
Compute loss and gradient for a minibatch of data.
Inputs:
- X: Array of input data of shape (N, d_1, ..., d_k)
- y: Array of labels, of shape (N,). y[i] gives the label for X[i].
Returns:
If y is None, then run a test-time forward pass of the model and return:
- scores: Array of shape (N, C) giving classification scores, where
scores[i, c] is the classification score for X[i] and class c.
If y is not None, then run a training-time forward and backward pass and
return a tuple of:
- loss: Scalar value giving the loss
- grads: Dictionary with the same keys as self.params, mapping parameter
names to gradients of the loss with respect to those parameters.
"""
scores = None
############################################################################
# TODO: Implement the forward pass for the two-layer net, computing the #
# class scores for X and storing them in the scores variable. #
############################################################################
# Unpack variables from the params dictionary
W1, b1 = self.params['W1'], self.params['b1']
W2, b2 = self.params['W2'], self.params['b2']
N, D = X.shape
# Compute the forward pass
scores = None
h1 = np.maximum(0, np.dot(X, W1) + b1) # 隐藏层末端有 relu函数
scores = np.dot(h1, W2) + b2
############################################################################
# END OF YOUR CODE #
############################################################################
# If y is None then we are in test mode so just return scores
if y is None:
return scores
# computer the loss and gradient
loss = None
scores = scores - np.reshape(np.max(scores, axis=1), (N, -1))
p = np.exp(scores) / np.reshape(np.sum(np.exp(scores), axis=1), (N, -1))
loss = -np.sum(np.log(p[range(N), list(y)])) / N
loss += 0.5 * reg * np.sum(W1 * W1) + 0.5 * reg * np.sum(W2 * W2)
# compute grads 这里的求导过程和之前的softmax是相似的，正确分类括号内减1 之后再3
grads = {}
dscores = p
dscores[range(N), list(y)] -= 1.0
dscores /= N
dw2 = np.dot(h1.T, dscores)
dh2 = np.sum(dscores, axis=0, keepdims=False)
da2 = np.dot(dscores, W2.T)
da2[h1 <= 0] = 0
dw1 = np.dot(X.T, da2)
dh1 = np.sum(da2, axis=0, keepdims=False)
dw2 += reg * W2
dw1 += reg * W1
grads['W1'] = dw1
grads['b1'] = dh1
grads['W2'] = dw2
grads['b2'] = dh2
############################################################################
# TODO: Implement the backward pass for the two-layer net. Store the loss #
# in the loss variable and gradients in the grads dictionary. Compute data #
# loss using softmax, and make sure that grads[k] holds the gradients for #
# self.params[k]. Don't forget to add L2 regularization! #
# #
# NOTE: To ensure that your implementation matches ours and you pass the #
# automated tests, make sure that your L2 regularization includes a factor #
# of 0.5 to simplify the expression for the gradient. #
############################################################################
############################################################################
# END OF YOUR CODE #
############################################################################
return loss, grads
def train(self,X,y,X_val,y_val,
learning_rate=1e-3,learning_rate_decay=0.95,
reg=1e-5,num_iters=100,
batch_size=200,verbose=True):
"""
train this neural network using SGD
inputs:
:param X: a numpy array of shape (N,D),giving training data
:param y: a numpy array of shape (N,),giving training labels;y[i] =c means that x[i] has label.where 0<=c< C
:param X_val: validation data,a numpy array of shape(N_val,D),
:param y_val: validation label,a numpy array of shape(N_val,)
:param learning_rate: 学习率
:param learning_rate_decay:Scalar giving factor used to decay the learning rate after each epoch.
:param reg:正则化的系数
:param num_iters: 迭代次数
:param batch_size: 每次执行的数据数量
:param verbose:boolean if true print progress during optimizing
:return:a dictionary{
'loss history': loss_history,
'train_acc_history':train_acc_history,
'val_acc_history':val_acc_history,
}
"""
num_train = X.shape[0]
iterations_per_epoch = max(num_train / batch_size, 1) #总的训练数据 除以 batch_size
# 这里要理清楚这里面的关系，每次迭代中，所有的训练数据都要进行运算一遍的，那么每次迭代中，一个batch_size是200，所以大迭代中还要进行多次运算
# 错错错，每次迭代，只用了batch_size大小的数据量来进行训练，这就是随机梯度法，通过一个较大的迭代次数，效果也行，而且计算代价降低了很多
# Use SGD to optimize the parameters in self.model
loss_history = []
train_acc_history =[]
val_acc_history =[]
for i in range(num_iters): # 这是最外层的大的迭代次数
X_batch = None
y_batch = None
indices = np.random.choice(num_train,batch_size)
X_batch = X[indices]
y_batch = y[indices]
#########################################################################
# TODO: Create a random minibatch of training data and labels, storing #
# them in X_batch and y_batch respectively. #
#########################################################################
#########################################################################
# END OF YOUR CODE #
#########################################################################
# Compute loss and gradients using the current minibatch
loss, grads = self.loss(X_batch,y_batch,reg=reg)
loss_history.append(loss)
W1 = grads['W1']
b1 = grads['b1']
W2 = grads['W2']
b2 = grads['b2']
self.params['W1'] -=learning_rate * W1
self.params['b1'] -= learning_rate * b1
self.params['W2'] -= learning_rate * W2
self.params['b2'] -= learning_rate * b2
#########################################################################
# TODO: Use the gradients in the grads dictionary to update the #
# parameters of the network (stored in the dictionary self.params) #
# using stochastic gradient descent. You'll need to use the gradients #
# stored in the grads dictionary defined above. #
#########################################################################
#########################################################################
# END OF YOUR CODE #
#########################################################################
if verbose and i % 100 == 0:
print('iteration %d / %d:loss %f' % (i,num_iters,loss))
# every epoch, check train and val accuracy and decay learning rate
if i % iterations_per_epoch ==0: #相当于minibatch经过多次迭代，数据量刚好达到训练数据量
train_acc = np.mean(self.predict(X_batch)== y_batch)
val_acc = np.mean(self.predict(X_val)==y_val)
train_acc_history.append(train_acc)
val_acc_history.append(val_acc)
# decay learning rate
learning_rate *= learning_rate_decay
return {
'loss_history':train_acc_history,
'train_acc_history':train_acc_history,
'val_acc_history':val_acc_history,
}
def predict(self,X):
"""
using trained weights of this two-layer network to predict labels for data points
For each data point we predict scores for each of the C classes, and assign each data point to the class
with the highest score.
:param X: a numpy array of shape(N,D) giving N D-dimensional data points to classify
:return: y_pred: A numpy array of shape (N,) giving predicted labels for each of
the elements of X. For all i, y_pred[i] = c means that X[i] is predicted
to have class c, where 0 <= c < C.
"""
# 用上一步训练出来的权重矩阵来计算分数并进行分类
y_pred = None
h1 = np.maximum(0,np.dot(X,self.params['W1'])+self.params['b1'])
scores = np.dot(h1,self.params['W2'])+self.params['b2']
y_pred = np.argmax(scores,axis=1)
###########################################################################
# TODO: Implement this function; it should be VERY simple! #
###########################################################################
###########################################################################
# END OF YOUR CODE #
###########################################################################
return y_pred
# input_dim=3 * 32 * 32, hidden_dim=100, num_classes=10,
# weight_scale=1e-3, reg=0.0
input_dim = 32*32*3
hidden_dim = 100
num_classes = 10
neural_network = TwoLayerNet(input_dim,hidden_dim,num_classes)
X_train,y_train,X_val,y_val,X_test,y_test = loadfile()
# train the network
stats = neural_network.train(X_train,y_train,X_val,y_val,num_iters=10000,
batch_size=200,learning_rate=1e-4,learning_rate_decay=0.95,
reg=0.5,verbose=True)
#predict on the validation set
plt.plot(stats['loss_history'])
plt.title("loss history")
plt.ylabel('Loss')
plt.show()
val_acc = np.mean(neural_network.predict(X_val)==y_val)
print("val accuracy:",val_acc)
test_acc = np.mean(neural_network.predict(X_test) == y_test)
print("test accuracy:",test_acc)
根据网上博主的代码自己跟着敲了一遍，我实在太菜了，上面的代码中还有一些步骤没有做，比如要进行超参数的测试验证啥的，而且这些超参数我是直接设定的比较好的结果，完全自己来的话当然要采取交叉验证的方式来去做，几个超参数混合搭配找到其中最好的组合。