準備

裏でTensorflowを使っているので、Tensorflowを入れる

pip install tensorflow

Kerasのインストール

pip install keras

上記サイトでモデルの構造を画像で保存する機能を使っているので、その準備

brew install graphviz
pip install pydot

実装の写経

ただ上記を真似しただけ。

# データセット準備
from keras.datasets import mnist
from keras.utils import np_utils

(x_train, y_train), (x_test, y_test) = mnist.load_data()

　#　インプットの形式変換
x_train = x_train.reshape(len(x_train),784)
x_test = x_test.reshape(len(x_test), 784)

x_train = x_train.astype("float32")
x_test = x_test.astype("float32")

x_train /= 255
x_test /= 255

print(x_train.shape[0], "train samples")
print(x_test.shape[0], "test samples")

nb_classes = 10

y_train = np_utils.to_categorical(y_train,nb_classes)
y_test = np_utils.to_categorical(y_test, nb_classes)


# モデル定義
from keras.models import Sequential
from keras.layers.core import Dense, Dropout, Activation
from keras.optimizers import Adam

def build_multilayer_perceptron():
    model = Sequential()
    model.add(Dense(512, input_shape=(784,)))
    model.add(Activation("relu"))
    model.add(Dropout(0.2))
    model.add(Dense(512))
    model.add(Activation("relu"))
    model.add(Dropout(0.2))
    model.add(Dense(10))
    model.add(Activation("softmax"))
    return model

model = build_multilayer_perceptron()

# modelのsummaryを表示＆画像でも表示
from keras.utils.vis_utils import plot_model

model.summary()
plot_model(model, show_shapes=True, show_layer_names=True, to_file="model.png")

# model構築
model.compile(loss="categorical_crossentropy", optimizer=Adam(), metrics=["accuracy"])

# 学習
from keras.callbacks import EarlyStopping

early_stopping = EarlyStopping(patience=0, verbose=1)
history = model.fit(x_train, y_train, batch_size=32, epochs=100, verbose=1, validation_split=0.1, callbacks=[early_stopping])

# 予測
model.evaluate(x_test, y_test)
# [0.07958280239470332, 0.979]

KerasでMNISTのメモ

準備

実装の写経