train_X, train_y = train[train_index], labels[train_index] test_X, test_y = train[test_index], labels[test_index] ################################################################################ ## list of prediction results from different classifiers predictions = [] ############################# ## XGBoost Classifier ############################# xgb_param = param.xgboost_para() xgboost_clf = clf.xgboost_classifer(xgb_param, train_X, train_y); ## xgboost input data requires specifc format xg_test = xgb.DMatrix(test_X, label=test_y) ## prediction from xgboost y_prob = xgboost_clf.predict(xg_test) predictions.append(y_prob) ############################# ## Lasagne NN Classifier #############################
for train_index, test_index in sss:
print 'split the training data'
train_X, train_y = train[train_index], labels[train_index]
test_X, test_y = train[test_index], labels[test_index]
################################################################################
## list of prediction results from different classifiers
predictions = []
#############################
## XGBoost Classifier
#############################
xgb_param = param.xgboost_para()
xgboost_clf = clf.xgboost_classifer(xgb_param, train_X, train_y)
## xgboost input data requires specifc format
xg_test = xgb.DMatrix(test_X, label=test_y)
## prediction from xgboost
y_prob = xgboost_clf.predict(xg_test)
predictions.append(y_prob)
#############################
## Lasagne NN Classifier
#############################
# standardize the data for NN