Question

我正在尝试创建一个keras CNN来预测我的ECG数据。我的x_train的形状像640,3498,1，y_train的形状像640,2（具有to_categorical 2类），我正在尝试构建模型。

X_train：

  [[[ 3.96889100e-12]
  [ 2.96721687e-12]
  [ 3.76118701e-12]
  ...
  [ 4.50912277e-12]
  [ 4.10963106e-12]
  [ 4.28550483e-12]]

 [[-3.28395026e-12]
  [-1.30043687e-12]
  [ 1.66892740e-12]
  ...
  [ 0.00000000e+00]
  [ 0.00000000e+00]
  [ 0.00000000e+00]]

 [[-3.71662156e-11]
  [-3.47206950e-11]
  [-3.72873674e-11]
  ...
  [-3.59975902e-11]
  [-3.60138075e-11]
  [-3.74937527e-11]]

y_train：

[[1. 0.]
 [0. 1.]
 [1. 0.]
 [1. 0.]
 [1. 0.]
 [1. 0.]
 [0. 1.]
 [0. 1.]
 [1. 0.]
 [0. 1.]]

我正在使用的代码是这样：

ctivation = 'relu'
model = Sequential()

model.add(Conv1D(filters=64, kernel_size=5, padding='same', kernel_regularizer=l2(0.01), bias_regularizer=l2(0.01),
                 kernel_initializer='normal', activation=activation, input_shape=X_train_dfm.shape[1:]))
model.add(MaxPool1D(pool_size=2)) 

model.add(Conv1D(filters=64, kernel_size=5, padding='same', kernel_regularizer=l2(0.01), bias_regularizer=l2(0.01),
                 kernel_initializer='normal', activation=activation))
model.add(MaxPool1D(pool_size=2)) 

model.add(Conv1D(filters=32, kernel_size=5, padding='same', kernel_regularizer=l2(0.01), bias_regularizer=l2(0.01),
                 kernel_initializer='normal', activation=activation))
model.add(MaxPool1D(pool_size=2)) 

model.add(Conv1D(filters=32, kernel_size=5, padding='same', kernel_regularizer=l2(0.01), bias_regularizer=l2(0.01),
                 kernel_initializer='normal', activation=activation))
model.add(MaxPool1D(pool_size=2))

model.add(Flatten())

model.add(Dense(2, activation='softmax'))

adam = keras.optimizers.Adam(lr=0.01)
model.compile(optimizer = adam, loss = 'categorical_crossentropy', metrics=['accuracy'])
model.summary()

期望值为整数0-1。但是，所有这些预测都是在〜-0.4和0.5之间浮动。

我尝试了10个测试数据。那就是我得到的

[[0.4986445  0.50135547]
 [0.4986445  0.50135547]
 [0.4986445  0.50135547]
 [0.4986445  0.50135547]
 [0.4986445  0.50135547]
 [0.4986445  0.50135547]
 [0.4986445  0.50135547]
 [0.4986445  0.50135547]
 [0.4986445  0.50135547]
 [0.4986445  0.50135547]]

val_loss, val_accuracy : [0.6938661336898804, 0.5062500238418579]

请帮助我，我一个星期就卡住了。任何帮助将不胜感激！

CNN Keras-意外的预测结果

0 个答案: