jddeguia/bagging-lstm

Implementation of bagging-based ensemble for solar irradiance prediction. Base learners used in ensemble learning is stacked-LSTM

bagging-lstm

Implementation of bagging-based ensemble for solar irradiance prediction. Base learners used in ensemble learning is stacked-LSTM

bagging-lstm's Issues

Issue in loading new data for training and testin

Dear Sir
I think you should change your code in the following part :
n_splits = 20
members = list()

for _ in range(n_splits):
model = evaluate_model(X_train, target, X_test, test_target)
members.append(model)
Because you only train the network with the same data.

Additionally, in the following part
single_scores, ensemble_scores = list(), list()
for i in range(1, n_splits+1):
subset = members[:i]
yhats = [model.predict(X_test) for model in subset]

you only evaluate the model with only one test data (model.predict(X_test))