上一篇说到了线性回归,它是一个回归的例子,那么本章主要介绍一个分类的例子(逻辑回归)。
如上图所示,分类模型与回归模型主要的区别就是输出节点从一个变成了多个。
PS:本章做分类所使用的数据集是Fashion-MNIST数据集(服饰类),它与mnist数据集非常像。
1、数据获取
mnist_train=gn.data.vision.FashionMNIST(train=True)
mnist_test=gn.data.vision.FashionMNIST(train=False)
data,label=mnist_train[0]
print(data.shape,label) # 查看数据维度
结果:
可以看到,该数据集是一张28X28的单通道图像,标签为2。打印图片看看结果:
def show_image(image): # 显示图像
n=image.shape[0]
_,figs=plt.subplots(1,n,figsize=(15,15))
for i in range(n):
figs[i].imshow(image[i].reshape((28,28)).asnumpy())
plt.show()
def get_fashion_mnist_labels(labels):# 显示图像标签
text_labels = ['t-shirt', 'trouser', 'pullover', 'dress', 'coat',
'sandal', 'shirt', 'sneaker', 'bag', 'ankle boot']
return [text_labels[int(i)] for i in labels]
show_image(data)
print(get_fashion_mnist_labels(label))
结果:
2、数据读取
batch_size=100
train_data=gn.data.DataLoader(dataset=mnist_train,batch_size=batch_size,shuffle=True)
test_data=gn.data.DataLoader(dataset=mnist_test,batch_size=batch_size,shuffle=False)
3、定义数据归一化操作
def transform(data,label):
return data.astype("float32")/255,label.astype("float32") # 样本归一化
4、初始化模型参数
num_input=28*28*1
num_output=10
w=nd.random_normal(shape=(num_input,num_output))
b=nd.random_normal(shape=(num_output))
params=[w,b]
与前几章一样,要对求导的参数开梯度:
for param in params:
param.attach_grad()
5、定义归一化函数
def softmax(x):
exp=nd.exp(x) # exp是一个矩阵
partition=exp.sum(axis=1,keepdims=True)
return exp/partition
6、定义网络
def net(X):
return softmax(nd.dot(X.reshape(-1, num_input), w) + b)
7、交叉熵损失函数
def cross_entropy(y_pre,y_true):
return -nd.pick(nd.log(y_pre),y_true)
8、定义准确率
def accuracy(output,label):
return nd.mean(output.argmax(axis=1)==label).asscalar()
9、定义测试集准确率
def evaluate_accuracy(data_iter,net):# 定义测试集准确率
acc=0
for data,label in data_iter:
data,label=transform(data,label)
output=net(data)
acc+=accuracy(output,label)
return acc/len(data_iter)
10、SGD梯度下降优化器
def SGD(params,lr):
for pa in params:
pa[:]=pa-lr*pa.grad # 参数沿着梯度的反方向走特定距离
11、训练
lr=0.1
epochs=20
for epoch in range(epochs):
train_loss=0
train_acc=0
for image,y in train_data:
image,y=transform(image,y) # 类型转换,数据归一化
with ag.record():
output=net(image)
loss=cross_entropy(output,y)
loss.backward()
# 将梯度做平均,这样学习率不会对batch_size那么敏感
SGD(params,lr/batch_size)
train_loss+=nd.mean(loss).asscalar()
train_acc+=accuracy(output,y)
test_acc=evaluate_accuracy(test_data,net)
print("Epoch %d, Loss:%f, Train acc:%f, Test acc:%f"
%(epoch,train_loss/len(train_data),train_acc/len(train_data),test_acc))
训练结果:
可以看到,loss值在慢慢减少,准确率均在提高。
12、预测
训练完成后,可对样本进行预测。
image_10,label_10=mnist_test[:10] #拿到前10个数据
show_image(image_10)
print("真实样本标签:",label_10)
print("真实数字标签对应的服饰名:",get_fashion_mnist_labels(label_10))
image_10,label_10=transform(image_10,label_10)
predict_label=net(image_10).argmax(axis=1)
print("预测样本标签:",predict_label.astype("int8"))
print("预测数字标签对应的服饰名:",get_fashion_mnist_labels(predict_label.asnumpy()))
预测结果:
所有源码:
import mxnet.gluon as gn
import mxnet.autograd as ag
import mxnet.ndarray as nd
def transform(data, label):
return data.astype("float32") / 255, label.astype("float32") # 样本归一化
mnist_train = gn.data.vision.FashionMNIST(train=True)
mnist_test = gn.data.vision.FashionMNIST(train=False)
data, label = mnist_train[0:9]
print(data.shape, label) # 查看数据维度
import matplotlib.pyplot as plt
def show_image(image): # 显示图像
n = image.shape[0]
_, figs = plt.subplots(1, n, figsize=(15, 15))
for i in range(n):
figs[i].imshow(image[i].reshape((28, 28)).asnumpy())
plt.show()
def get_fashion_mnist_labels(labels): # 显示图像标签
text_labels = ['t-shirt', 'trouser', 'pullover', 'dress', 'coat',
'sandal', 'shirt', 'sneaker', 'bag', 'ankle boot']
return [text_labels[int(i)] for i in labels]
#
# show_image(data)
# print(get_fashion_mnist_labels(label))
'''----数据读取----'''
batch_size = 100
transformer = gn.data.vision.transforms.ToTensor()
train_data = gn.data.DataLoader(dataset=mnist_train, batch_size=batch_size, shuffle=True)
test_data = gn.data.DataLoader(dataset=mnist_test, batch_size=batch_size, shuffle=False)
'''----初始化模型参数----'''
num_input = 28 * 28 * 1
num_output = 10
w = nd.random_normal(shape=(num_input, num_output))
b = nd.random_normal(shape=(num_output))
params = [w, b]
for param in params:
param.attach_grad()
# 定义损失函数
def softmax(x):
exp = nd.exp(x) # exp是一个矩阵
partition = exp.sum(axis=1, keepdims=True)
return exp / partition
# 定义模型
def net(X):
return softmax(nd.dot(X.reshape(-1, num_input), w) + b)
# 交叉熵损失函数
def cross_entropy(y_pre,y_true):
return -nd.pick(nd.log(y_pre),y_true)
# 定义准确率
def accuracy(output,label):
return nd.mean(output.argmax(axis=1)==label).asscalar()
def evaluate_accuracy(data_iter,net):# 定义测试集准确率
acc=0
for data,label in data_iter:
data,label=transform(data,label)
output=net(data)
acc+=accuracy(output,label)
return acc/len(data_iter)
# 梯度下降优化器
def SGD(params,lr):
for pa in params:
pa[:]=pa-lr*pa.grad # 参数沿着梯度的反方向走特定距离
'''---训练---'''
lr=0.1
epochs=20
for epoch in range(epochs):
train_loss=0
train_acc=0
for image,y in train_data:
image,y=transform(image,y) # 类型转换,数据归一化
with ag.record():
output=net(image)
loss=cross_entropy(output,y)
loss.backward()
# 将梯度做平均,这样学习率不会对batch_size那么敏感
SGD(params,lr/batch_size)
train_loss+=nd.mean(loss).asscalar()
train_acc+=accuracy(output,y)
test_acc=evaluate_accuracy(test_data,net)
print("Epoch %d, Loss:%f, Train acc:%f, Test acc:%f"
%(epoch,train_loss/len(train_data),train_acc/len(train_data),test_acc))
'''----预测-------'''
# 训练完成后,可对样本进行预测
image_10,label_10=mnist_test[:10] #拿到前10个数据
show_image(image_10)
print("真实样本标签:",label_10)
print("真实数字标签对应的服饰名:",get_fashion_mnist_labels(label_10))
image_10,label_10=transform(image_10,label_10)
predict_label=net(image_10).argmax(axis=1)
print("预测样本标签:",predict_label.astype("int8"))
print("预测数字标签对应的服饰名:",get_fashion_mnist_labels(predict_label.asnumpy()))