tf2 fashion_mnist 入门

学习使用tf2

视频教程传送门

知识点：

loss="sparse_categorical_crossentropy"

这个 sparse是对y进行one-hot操作，如果y已经做过one-hot，则使用 categorical_crossentropy.

#!/usr/bin/env python
# coding: utf-8

# In[1]:


import tensorflow as tf
import tensorflow.keras as k
import numpy as np
import matplotlib.pyplot as plt


# In[21]:


fashion_mnist = k.datasets.fashion_mnist
(x_train,y_train),(x_test,y_test)=fashion_mnist.load_data()
x_train,x_valid = x_train[:5000],x_train[5000:]
y_train,y_valid= y_train[:5000],y_train[5000:]


# In[7]:


def show_single_img(img):
    plt.imshow(img,cmap="binary")
    plt.show()


# In[8]:


show_single_img(x_vaild[0])


# In[16]:


def show_imgs(n_rows,n_cols,x,y,classes):
    plt.figure(figsize=(n_rows*1.4,n_cols*1.6))
    for row in range(n_rows):
        for col in range(n_cols):
            index = row * n_cols + col
            plt.subplot(n_rows,n_cols,index+1)
            plt.imshow(x[index],cmap="binary")
            plt.title(classes[y[index]])
            plt.axis("off")
classes=['T-shirt/top','Trouser','Pullover','Dress','Coat',
         'Sandal','Shirt','Sneaker','Bag','Ankle boot']


# In[17]:


show_imgs(1,5,x_train[:5],y_train[:5],classes)


# In[24]:


#build the model
model =k.Sequential()
model.add(k.layers.Flatten(input_shape=[28,28]))
model.add(k.layers.Dense(300,activation="relu"))
model.add(k.layers.Dense(100,activation="relu"))
model.add(k.layers.Dense(10,activation="softmax"))

model.compile(loss="sparse_categorical_crossentropy",
             optimizer="adam",
             metrics=["accuracy"])


# In[25]:


history=model.fit(x_train,y_train,epochs=10,
         validation_data=(x_valid,y_valid))


# In[27]:


import pandas as pd
def plot_curve(history):
    pd.DataFrame(history.history).plot(figsize=(8,5))
    plt.grid(True)
    plt.gca().set_ylim(0,1)
    plt.show()
plot_curve(history)


# In[ ]:

适用sklearn对数据集进行归一化操作

#!/usr/bin/env python
# coding: utf-8

# In[5]:


import tensorflow as tf
import tensorflow.keras as k
import numpy as np
import matplotlib.pyplot as plt


# In[6]:


fashion_mnist = k.datasets.fashion_mnist
(x_train,y_train),(x_test,y_test)=fashion_mnist.load_data()
x_train,x_valid = x_train[:5000],x_train[5000:]
y_train,y_valid= y_train[:5000],y_train[5000:]


# In[7]:


from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
x_train_scaled = scaler.fit_transform(x_train.astype(np.float32).reshape(-1,1)).reshape(-1,28,28)
x_valid_scaled = scaler.transform(x_valid.astype(np.float32).reshape(-1,1)).reshape(-1,28,28)
x_test_scaled = scaler.transform(x_test.astype(np.float32).reshape(-1,1)).reshape(-1,28,28)


# In[8]:


#build the model
model =k.Sequential()
model.add(k.layers.Flatten(input_shape=[28,28]))
model.add(k.layers.Dense(300,activation="relu"))
model.add(k.layers.Dense(100,activation="relu"))
model.add(k.layers.Dense(10,activation="softmax"))

model.compile(loss="sparse_categorical_crossentropy",
             optimizer="adam",
             metrics=["accuracy"])


# In[9]:


history=model.fit(x_train,y_train,epochs=10,
         validation_data=(x_valid,y_valid))


# In[10]:


import pandas as pd
def plot_curve(history):
    pd.DataFrame(history.history).plot(figsize=(8,5))
    plt.grid(True)
    plt.gca().set_ylim(0,1)
    plt.show()
plot_curve(history)


# In[ ]: