• • 已关闭。**此问题为not reproducible or was caused by typos。当前不接受答案。

这个问题是由打字错误或无法再重现的问题引起的。虽然类似的问题在这里可能是on-topic，但这个问题的解决方式不太可能帮助未来的读者。
3小时前关门了。
Improve this question
我很感激你花时间考虑我的问题。
我有一个问题与lstm，因为它给了我overfiting为我所有的结果。我尝试了不同的技术。
有人可以检查我的代码，看看我是否写错了行吗？

## load input stages
dfList = []
rowsCounts=[]
for i in range(1,41):
    rowsCounts.append(pd.read_csv(r'C:\Users\wannees.alkhayyali\OneDrive - University of Calgary\PFPD\Field Data\COP\inj data\stage_'+str(i)+'.csv',usecols = [i for i in range(10)]).shape[0])
maxRowsCount=max(rowsCounts)

# Fill rows with zero to ensure that each stage has the same length

for i in range(1,41):
    df=pd.read_csv(r'C:\Users\wannees.alkhayyali\OneDrive - University of Calgary\PFPD\Field Data\COP\inj data\stage_'+str(i)+'.csv',usecols = ['stage_clean_cum_m3','clean_rate_m3_per_min','proppant_con (kg/m3)'])
    df_final = df.reindex(range(maxRowsCount), fill_value=0)
  
# Normalize
    min_max_scaler = preprocessing.MinMaxScaler()
    x_scaled = min_max_scaler.fit_transform(df_final.values)
    dfList.append(x_scaled)
dataSet = np.array(dfList)

dfOutput=pd.read_csv(r'C:\Users\wannees.alkhayyali\OneDrive - University of Calgary\PFPD\Field Data\COP\inj data\output.csv')
Y = np.array(dfOutput.values)

#Split into train and test sets
Xtrain=dataSet[0:30]
Xtest=dataSet[30:]
Ytrain=Y[0:30]
Ytest=Y[30:]

#Design network
model = Sequential()
model.add(LSTM(50, return_sequences=True, kernel_regularizer=l2(0.01), recurrent_regularizer=l2(0.01), bias_regularizer=l2(0.01), input_shape=(maxRowsCount, 3)))
model.add(LSTM(20))
# model.add(LSTM(50, activation='relu'))
model.add(Dense(1, activation='sigmoid'))
opt= optimizers.Adam(learning_rate=0.0001)
model.compile(optimizer=opt, loss='mse')