调用图层“sequential”（类型 Sequential）时遇到异常

Question

我正在尝试执行这个包含 6 个输入和 6 个输出的神经网络，但抛出了异常

ValueError: Exception encountered when calling layer 'sequential' (type Sequential).

Input 0 of layer "dense" is incompatible with the layer: expected min_ndim=2, found ndim=1. Full shape received: (None,)

# Entradas
x1 = 0
x2 = 0
x3 = 0
x4 = 0
x5 = 0
x6 = 0

# Salidas
y1 = 0
y2 = 0
y3 = 0
y4 = 0
y5 = 0
y6 = 1

# crea el modelo
model = Sequential()
model.add(Dense(12, input_dim=6, activation='relu'))
model.add(Dense(6, activation='sigmoid'))


# Compila el modelo
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])

# Ajusta el modelo
model.fit([x1,x2,x3,x4,x5,x6], [y1,y2,y3,y4,y5,y6], epochs=150, batch_size=10)


# evalua el modelo
scores = model.evaluate([x1,x2,x3,x4,x5,x6], [y1,y2,y3,y4,y5,y6])

print("\n%s: %.2f%%" % (model.metrics_names[1], scores[1]*100))

Answer 1

您的输入和 output 尺寸应为：

(number_of_data_input,input_dimension) 
(number_of_data_output,output_dimension)

由于您有 1 个输入和 1 个 output 都是 6 维。 因此，您应该在 fit 中传递(1,6)维度并评估函数而不是(6,) 。 所以适合 function 应该是：

model.fit([[x1,x2,x3,x4,x5,x6]], [[y1,y2,y3,y4,y5,y6]], epochs=150, batch_size=10)

并评估 function 应该是：

scores = model.evaluate([[x1,x2,x3,x4,x5,x6]], [[y1,y2,y3,y4,y5,y6]])