LSTM 的输入形状，ValueError：“sequential_9”层的输入 0 与该层不兼容：预期形状=(None, 1, 13)，找到的形状=(None, 13)

Question

我正在尝试解决分类问题，但我不知道为什么会出现此错误：

ValueError: Input 0 of layer "sequential_9" is incompatible with the layer: expected shape=(None, 1, 13), found shape=(None, 13)

这是我的代码：

import pandas as pd
#import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.ensemble import VotingClassifier
from imblearn.over_sampling import RandomOverSampler
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, roc_auc_score, confusion_matrix, classification_report
import time as t
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv1D, MaxPooling1D, Flatten, Dense, LSTM, Dropout, Input
from tensorflow.keras.wrappers.scikit_learn import KerasClassifier


data = pd.read_csv('Dataset.csv')
print(data.dtypes)

X = data.drop('class', axis=1)
y = data['class']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

start_time = t.time()
print("Starting oversampling process :")
# Implement oversampling to balance classes
oversample = RandomOverSampler(sampling_strategy='minority', random_state=42)
X_train, y_train = oversample.fit_resample(X_train, y_train)
print("FINAL: oversampling process. Time to done process : ",(t.time()-start_time))

def create_model():
    model = Sequential()
    model.add(Conv1D(32, kernel_size=3, activation='relu', input_shape=(X_train.shape[1], 1)))
    model.add(Conv1D(64, kernel_size=3, activation='relu'))
    model.add(MaxPooling1D(pool_size=2))
    model.add(Flatten())
    model.add(Dense(128, activation='relu'))
    model.add(Dense(1, activation='sigmoid'))
    model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
    return model

cnn_model = KerasClassifier(build_fn=create_model, epochs=20, batch_size=32, verbose=0)
cnn_model._estimator_type = "classifier"

# Define the LSTM model
def create_LSTM_model():
    model = Sequential()
    model.add(LSTM(128, input_shape=(1,13), return_sequences=True))
    model.add(Dense(32, activation='relu'))
    model.add(Dropout(0.2))
    model.add(LSTM(64))
    model.add(Dropout(0.2))
    model.add(Dense(32, activation='relu'))
    model.add(Dense(2, activation='sigmoid'))
    model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
    return model

lstm_model = KerasClassifier(build_fn=create_LSTM_model, epochs=20, batch_size=32, verbose=0)
lstm_model._estimator_type = "classifier"



# define the ensemble model
voting_clf = VotingClassifier(estimators=[("CNN", cnn_model), ("LSTM", lstm_model)], voting='soft')
voting_clf.fit(X_train, y_train)


print("Evaluate the ensemble model on the test data : ")
ensemble_pred = voting_clf.predict(X_test)
ensemble_accuracy = accuracy_score(y_test, ensemble_pred)
ensemble_precision = precision_score(y_test, ensemble_pred)
ensemble_recall = recall_score(y_test, ensemble_pred)
ensemble_f1_score = f1_score(y_test, ensemble_pred)
ensemble_auc_roc = roc_auc_score(y_test, ensemble_pred)
ensemble_cm = confusion_matrix(y_test, ensemble_pred)
CR = classification_report(y_test, ensemble_pred)

总的来说，我有以下 14 个特征：

f1                        int64
f2                        int64
f3                        int64
f4                      float64
f5                        int64
f6                        int64
f7                      float64
f8                      float64
f9                        int64
f10                       int64
f11                       int64
f12                       int64
f13                       int64
class                     int64
dtype: object

和

X_train.shape

(11725212, 13)

当我使用

input_shape=(X_train.shape[1], 1)

时它有效，但我认为这是错误的。 所以，我尝试了

input_shape=(1,13)

和

input_shape=(13,)

但仍然出现错误。

有人可以帮忙吗？谢谢