def evaluate_smm(player_data, id):
models = [LogisticRegression(penalty='l1'),
MultinomialNB(),
DecisionTreeClassifier(),
SVC(probability=True, kernel='rbf', gamma=0.3),
RandomForestClassifier(n_estimators=20, n_jobs=-1)]
player_model = StackingModel(models)
results = validate.cross_validate(player_data, player_model, 10)
validate.save_roc_curves(results, "smm/" + str(id))
def evaluate_forest(player_data, id):
player_model = RandomForestClassifier(n_estimators=20, n_jobs=-1)
results = validate.cross_validate(player_data, player_model, 10)
validate.save_roc_curves(results, "forest/" + str(id))
def evaluate_svm(player_data, id):
player_model = SVC(probability=True, kernel='rbf', gamma=0.3)
results = validate.cross_validate(player_data, player_model, 10)
validate.save_roc_curves(results, "svm/" + str(id))
def evaluate_dtree(player_data, id):
player_model = DecisionTreeClassifier()
results = validate.cross_validate(player_data, player_model, 10)
validate.save_roc_curves(results, "dtree/" + str(id))
def evaluate_naive(player_data, id):
player_model = MultinomialNB()
results = validate.cross_validate(player_data, player_model, 10)
validate.save_roc_curves(results, "naive/" + str(id))
def evaluate_logit(player_data, id):
player_model = LogisticRegression(penalty='l1')
results = validate.cross_validate(player_data, player_model, 10)
validate.save_roc_curves(results, "logit/" + str(id))