Kerasを用いたCNNの機械学習のサンプル数の不一致

Question

CNNをもとに機械学習を行っているのですが、サンプル数は683と出ているのですが、機械学習の際に使われている数が22となっています。

この原因が分かる方がいらっしゃれば是非教えていただきたいです。

該当のソースコード

CNNを構築

model = Sequential()

model.add(Conv2D(32, (3, 3), padding='same',input_shape=X_train.shape[1:]))
model.add(Activation('relu'))
model.add(Conv2D(32, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))

model.add(Conv2D(64, (3, 3), padding='same'))
model.add(Activation('relu'))
model.add(Conv2D(64, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))

model.add(Flatten())
model.add(Dense(512))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(3))
model.add(Activation('softmax'))

model.summary()

コンパイル

model.compile(loss='categorical_crossentropy',optimizer='SGD',metrics=['accuracy'])
print(len(X_train))
print(len(y_train))
print(len(X_test))
print(len(y_test))
history = model.fit(X_train, y_train, epochs=5)

#出力結果

Model: "sequential"
Layer (type) Output Shape Param #

conv2d (Conv2D) (None, 100, 100, 32) 896

---

activation (Activation) (None, 100, 100, 32) 0

---

conv2d_1 (Conv2D) (None, 98, 98, 32) 9248

---

activation_1 (Activation) (None, 98, 98, 32) 0

---

max_pooling2d (MaxPooling2D) (None, 49, 49, 32) 0

---

dropout (Dropout) (None, 49, 49, 32) 0

---

conv2d_2 (Conv2D) (None, 49, 49, 64) 18496

---

activation_2 (Activation) (None, 49, 49, 64) 0

---

conv2d_3 (Conv2D) (None, 47, 47, 64) 36928

---

activation_3 (Activation) (None, 47, 47, 64) 0

---

max_pooling2d_1 (MaxPooling2 (None, 23, 23, 64) 0

---

dropout_1 (Dropout) (None, 23, 23, 64) 0

---

flatten (Flatten) (None, 33856) 0

---

dense (Dense) (None, 512) 17334784

---

activation_4 (Activation) (None, 512) 0

---

dropout_2 (Dropout) (None, 512) 0

---

dense_1 (Dense) (None, 3) 1539

---

=================================================================
Total params: 17,401,891
Trainable params: 17,401,891
Non-trainable params: 0

---

683
683
171
171
Epoch 1/5
22/22 [==============================] - 29s 1s/step - loss: 1.1049 - accuracy: 0.3704
Epoch 2/5
22/22 [==============================] - 28s 1s/step - loss: 1.0947 - accuracy: 0.3953
Epoch 3/5
22/22 [==============================] - 29s 1s/step - loss: 1.0884 - accuracy: 0.4158
Epoch 4/5
22/22 [==============================] - 31s 1s/step - loss: 1.0827 - accuracy: 0.4085
Epoch 5/5
22/22 [==============================] - 40s 2s/step - loss: 1.0843 - accuracy: 0.4114

Answer 1

22というのはバッチ回数です。
サンプル数 683 ÷ バッチサイズデフォルト32 = 21.34375 で切り上げて22回
という計算になります。

Comments

[gunnar]

ありがとうございます。結果の見方が分からなかったのでとても参考になります