scikit-learn中的精确_score与Keras的精度之间的差异

Question

我已经在keras上实施并训练了多类卷积神经网络。最终的测试精度为0.9522。但是，当我使用Scikit-Learn的准确性_score计算准确性时，我获得了0.6224。这是我所做的：

X_train = X[:60000, :, :, :]
X_test = X[60000:, :, :, :]
y_train = y[:60000, :]
y_test = y[60000:, :]
print ('Size of the arrays:')
print ('X_train: ' + str(X_train.shape))
print ('X_test: ' + str(X_test.shape))
print ('y_train: ' + str(y_train.shape))
print ('y_test: ' + str(y_test.shape))

结果：

Size of the arrays:
X_train: (60000, 64, 64, 3)
X_test: (40000, 64, 64, 3)
y_train: (60000, 14)
y_test: (40000, 14)

拟合KERAS模型（我在这里不添加整个模型以使代码保持简单）：

model = Sequential()
model.add(Conv2D(10, (5,5), padding='same', input_shape=(64, 64, 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(14))
model.add(Activation('softmax'))
model.compile(optimizer='rmsprop', loss='mean_squared_error', metrics['accuracy'])
model.fit(X_train, y_train, batch_size=100, epochs=5, verbose=1, validation_data=(X_test, y_test))

Scikit-learn的准确性：

y_pred = model.predict(X_test, batch_size=100)
y_pred1D = y_pred.argmax(1)
y_pred = model.predict(X_test, batch_size=100)
y_test1D = y_test.argmax(1)
print ('Accuracy on validation data: ' + str(accuracy_score(y_test1D, y_pred1D)))

得分：

Accuracy on validation data: 0.6224

keras的准确性：

score_Keras = model.evaluate(X_test, y_test, batch_size=200)
print('Accuracy on validation data with Keras: ' + str(score_Keras[1]))

结果：

Accuracy on validation data with Keras: 0.95219109267

我的问题是：为什么两个精度有所不同，我应该使用哪一个来评估我的多类分类器的性能？

预先感谢！

Answer 1

您的代码中有一个错字，为什么要两次定义y_pred？

y_pred = model.predict(X_test, batch_size=100)
y_pred1D = y_pred.argmax(1)
y_pred = model.predict(X_test, batch_size=100)
y_test1D = y_test.argmax(1)
print ('Accuracy on validation data: ' + str(accuracy_score(y_test1D, y_pred1D)))

应该是：

y_pred = model.predict(X_test, batch_size=100)
y_pred1D = y_pred.argmax(1)
y_test1D = y_test.argmax(1)
print ('Accuracy on validation data: ' + str(accuracy_score(y_test1D, y_pred1D)))

尽管如此，您应该提供y_pred1D和y_test1D的值和形状，但是当您进行y_pred1D = y_pred.argmax(1)和y_test1D = y_test.argmax(1)时，错误在于错误，以便使用Scikit Learn Learn Metric。我的猜测是，这不是您认为的，否则两个指标将相同。