data_path = "data/"
X_train, X_test, Y_train, Y_test = get_all_data(data_path)
"""
Define model
"""
# model = RandomForestClassifier()
model = RandomForestClassifier(n_estimators=800, oob_score='TRUE')
"""
Crossvalidation
"""
detect_with_cross_validation(model, X_train, Y_train)
"""
Detecting
"""
model.fit(X_train, Y_train)
detect(model, X_test, Y_test)
# param = {
# 'n_estimators': [500, 600, 700, 800, 900, 1000, 1100, 1200, 1300, 1400]
# }
# kflod = StratifiedKFold(n_splits=10)
# grid_search = GridSearchCV(estimator=model, param_grid=param, scoring='accuracy', cv=kflod)
# grid_result = grid_search.fit(X_train, Y_train)
# print("Best: %f using %s" % (grid_result.best_score_, grid_search.best_params_))
#
# means = grid_result.cv_results_['mean_test_score']
# params = grid_result.cv_results_['params']
# for mean, param in zip(means,params):
# print("%f with: %r" % (mean, param))
pyplot.bar(range(len(model.feature_importances_)), model.feature_importances_)
""" Crossvalidation """ detect_with_cross_validation(clf, X_train, y_train) """ Learning """ clf.fit(X_train, y_train) """ Detect """ detect(clf, X_test, y_test) # score = clf.score(X_test, y_test) # print score # # iterate over classifiers # for clf in classifiers: # # ax = plt.subplot(len(datasets), len(classifiers) + 1, i) # clf.fit(X_train, y_train) # score = clf.score(X_test, y_test) # # clf.fit(X, y) # # score = clf.score(X_test_, y_test_) # print score
# subsample=0.8,
# colsample_bytree=0.8,
# objective='binary:logistic',
# nthread=4,
# scale_pos_weight=1,
# seed=27)
"""
Crossvalidation
"""
detect_with_cross_validation(model, np_X_train, np_y_train)
"""
Detect model
"""
model.fit(np_X_train, np_y_train)
detect(model, np_X_test, np_y_test)
# pyplot.bar(range(len(model.feature_importances_)), model.feature_importances_)
# pyplot.show()
# param = {
# 'n_estimators': [1300, 1400, 1500, 1600, 1700, 1800], #[600, 700, 800, 900, 950, 1000, 1050, 1100, 1150, 1200],
# 'learning_rate': [0.1],
# 'max_depth': [10],
# 'min_child_weight': [1],
# 'gamma': [0],
# 'subsample': [0.8],
# 'colsample_bytree': [0.8],
# 'objective': ['binary:logistic'],
# 'nthread': [4],
# 'scale_pos_weight': [1],