Dense(500, input_dim=4, kernel_initializer='normal',
activation='relu'))
model.add(Dense(1))
# Compile model
model.compile(loss='mean_absolute_percentage_error', optimizer='adam')
return model
dnn = make_pipeline(
RobustScaler(),
KerasRegressor(build_fn=NeuralNetModel,
epochs=10,
batch_size=200,
verbose=0))
stacked_averaged_models = StackingAveragedModels(
base_models=[model_xgb, model_lgb], meta_model=lasso, n_folds=4)
ensemble_reg = EnsembleRegressor([model_lgb])
#Mapes and Submission
mapes = []
tcv_mapes = []
cv_mapes = []
Sub = []
tcv = False
cv = False
val = False
#Cross-Val and Training
#Filter on the right town and keep the test index
train_town = df_train
def NeuralNetModel():
# create model
model = Sequential()
model.add(Dense(500, input_dim=4, kernel_initializer='normal', activation='relu'))
model.add(Dense(1))
# Compile model
model.compile(loss='mean_absolute_percentage_error', optimizer='adam')
return model
dnn = make_pipeline(PartScaler(last_col = 5), KerasRegressor(build_fn=NeuralNetModel, epochs=10, batch_size=200, verbose=0))
rfr = RandomForestRegressor(n_estimators = 10, n_jobs = -1, random_state = 123)
stacked_averaged_models = StackingAveragedModels(base_models = [model_slgb],
meta_model = model_sxgb,
n_folds = 4)
ensemble_reg = EnsembleRegressor([model_lgb])
#Mapes and Submission
mapes = []
tcv_mapes = []
cv_mapes = []
Sub = []
tcv = False
cv = False
val = False