3.keras-簡單實現Mnist數據集分類

keras-簡單實現Mnist數據集分類

1.載入數據以及預處理

import numpy as np
from keras.datasets import mnist
from keras.utils import np_utils
from keras.models import Sequential
from keras.layers import *
from keras.optimizers import SGD

import os

import tensorflow as tf

# 載入數據
(x_train,y_train),(x_test,y_test) = mnist.load_data()

# 預處理
# 將（60000，28，28）轉化爲（600000，784），好輸入展開層
x_train = x_train.reshape(x_train.shape[0],-1)/255.0
x_test= x_test.reshape(x_test.shape[0],-1)/255.0
# 將輸出轉化爲one_hot編碼
y_train = np_utils.to_categorical(y_train,num_classes=10)
y_test = np_utils.to_categorical(y_test,num_classes=10)

2.建立網絡打印訓練結果

# 建立網絡
model = Sequential([
    # 輸入784輸出10個
    Dense(units=10,input_dim=784,bias_initializer='one',activation='softmax')
])
# 編譯
# 自定義優化器
sgd = SGD(lr=0.1)
model.compile(optimizer=sgd,
              loss='mse',
              # 獲得訓練過程當中的準確率
              metrics=['accuracy'])

model.fit(x_train,y_train,batch_size=32,epochs=10,validation_split=0.2)

# 評估模型
loss,acc = model.evaluate(x_test,y_test,)
print('\ntest loss',loss)
print('test acc',acc)

out:

Epoch 1/10

32/48000 [..............................] - ETA: 2:27 - loss: 0.0905 - acc: 0.1875
1248/48000 [..............................] - ETA: 5s - loss: 0.0907 - acc: 0.1346

......

......

Epoch 10/10

45952/48000 [===========================>..] - ETA: 0s - loss: 0.0164 - acc: 0.9005
47616/48000 [============================>.] - ETA: 0s - loss: 0.0163 - acc: 0.9008
48000/48000 [==============================] - 2s 37us/step - loss: 0.0163 - acc: 0.9010 - val_loss: 0.0149 - val_acc: 0.9084

 

32/10000 [..............................] - ETA: 4s
3360/10000 [=========>....................] - ETA: 0s
5824/10000 [================>.............] - ETA: 0s
8512/10000 [========================>.....] - ETA: 0s
10000/10000 [==============================] - 0s 20us/step

test loss 0.015059704356454312test acc 0.908