TestModels.py

import CreateDataframe as data
from sklearn.ensemble import AdaBoostClassifier
from sklearn.ensemble import ExtraTreesClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.neural_network import MLPClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import BaggingClassifier
from sklearn.ensemble import GradientBoostingClassifier

# Hyper parameters
BOOSTER_ESTIMATOR_NUM = 200
TREES_ESTIMATOR_NUM = 100
TREES_MAX_DEPTH = 20
FOREST_ESTIMATOR_NUM = 100
FOREST_MAX_DEPTH = 20
K = 173
HIDDEN_LAYERS = 100

# data
train_features = data.train_p_features
train_labels = data.train_p_labels
validation_features = data.validation_p_features
validation_labels = data.validation_p_labels

# create models
booster = AdaBoostClassifier(n_estimators=BOOSTER_ESTIMATOR_NUM)
trees = ExtraTreesClassifier(n_estimators=TREES_ESTIMATOR_NUM, max_depth=TREES_MAX_DEPTH)
forest = RandomForestClassifier(n_estimators=FOREST_ESTIMATOR_NUM, max_depth=FOREST_MAX_DEPTH)
KNN = KNeighborsClassifier(n_neighbors=K)
logistic = LogisticRegression()
MLP = MLPClassifier(hidden_layer_sizes=HIDDEN_LAYERS)
bagging = BaggingClassifier()
gradient_boosting = GradientBoostingClassifier()

models = [booster, trees, forest, KNN, logistic, MLP, bagging, gradient_boosting]
models_names = ["Adaboost", "Extra Trees", "Random Forest", "KNN", "Logistic",
                "MLP", "Bagging", "Gradient Boosting"]

# train models
for model in models:
    model.fit(train_features, train_labels)

# print predicted accuracy on validation data
for i, model in enumerate(models):
    accuracy = "{:.3f}".format(model.score(validation_features, validation_labels))
    print("Accuracy for " + models_names[i] + " Classifier: " + str(accuracy))