Build Deep Learning model for voting classification

Question

I want to build the Voting classification model for the following data as I have two different approaches and apply to vote for the same. The data are presented here.

labels  New_data
0   0   [72.5, 83.5, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
1   0   [61.5, 55.5, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
2   0   [34.5, 38.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
3   0   [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
4   0   [0.0, 17.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]

Here New_data is considered as an X_train data and labels should be 0 or 1 and considered as y_train data.

Another data for X_train is presented below and y_train is the same as the previous value 0 and 1.

array([[ 17,  37,  19, ...,   1, 167,   0],
       [  4, 133, 103, ...,   1,  58,   8],
       [  6,  18, 197, ...,   0,   0,   0],
       ...,
       [ 75,   8,  17, ...,   6,  67,   6],
       [155,  25,  11, ...,   0,   0,   0],
       [  2,   1,  23, ..., 223,  53,  54]], dtype=int32)

The DL model Architecture for both approach is given below.

Approach 1:

def create_model():
  imp = keras.Input(shape=(len(data11["New_data"][0]), 1),dtype='int32')
  x = LSTM(64, return_sequences=True,name='lstm_layer')(imp)
  x = Conv1D(filters=128, kernel_size=3, padding='same', activation='relu')(x)
  x = MaxPooling1D(pool_size=2)(x)
  x = Dropout(0.1)(x)
  x = Flatten() (x)
  x = Dropout(0.1)(x)
  x= Dense(250, activation='sigmoid')(x) 
  x= Dense(250, activation='sigmoid')(x) 
  preds = tf.keras.layers.Dense(1, activation='sigmoid')(x)
  model = Model(inputs= imp, outputs = preds)
  return model

NN_clf=KerasClassifier(build_fn = lambda: create_model(), epochs = 15, batch_size = 32, verbose = 0)
NN_clf._estimator_type = "classifier"

Approach 2:

def Create_model11():
    embedding_vecor_length = 100
    model1 = Sequential()
    model1.add(Embedding(vocab_size, 32, input_length=maxlen))
    model1.add(LSTM(32, return_sequences=True))
    model1.add(Dense(1, activation='sigmoid'))
    model1.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
    model1.summary()
    return model1
  
     

 
  NN=KerasClassifier(build_fn= lambda: Create_model11(), epochs = 15, batch_size = 32)
  NN._estimator_type = "classifier"

I want to apply voting classification but am not able to perform this task can anyone please help me out? The voting classification code is written below.

voting_clf = VotingClassifier(estimators=[('Fusion', NN_clf), ('Normal', NN)])
voting_clf.fit(X_train, y_train)

Getting below error.

TypeError: Exception encountered when calling layer "lstm_layer" (type LSTM).

Input 'b' of 'MatMul' Op has type float32 that does not match type int32 of argument 'a'.

Call arguments received:
  • inputs=tf.Tensor(shape=(None, 8, 1), dtype=int32)
  • mask=None
  • training=None
  • initial_state=None

Build Deep Learning model for voting classification

0 Answers