def Classifier_LogisticRegression(X, y, resp1):
nome = "Logistic_Regression"
X_new, nomeFeature = selecao_feature(X, y, resp1)
# split data into train and test sets
seed = 100
test_size = 0.2
# separação treino - teste: 80 - 20
X_train, X_test, y_train, y_test = train_test_split(X_new,
y,
test_size=test_size,
random_state=seed)
# fit model no training data
## parametros do Modelo: https://xgboost.readthedocs.io/en/latest/python/python_api.html
model = LogisticRegression()
model.fit(X_train, y_train)
# make predictions for test data
start_time = utils.get_time() # Tempo inicial
y_pred = model.predict(X_test)
predictions = [round(value) for value in y_pred]
diff_time = utils.get_time_diff(start_time) # Tempo final
print_resultados(model, y_test, predictions, diff_time, X_test, nome,
nomeFeature)
def Classifier_Gradient_Boosting(X, y, resp1):
nome = "Gradient_Boosting"
X_new, nomeFeature = selecao_feature(X, y, resp1)
# split data into train and test sets
seed = 0
test_size = 0.2
n_estimators = 200
# separação treino - teste: 80 - 20
X_train, X_test, y_train, y_test = train_test_split(X_new,
y,
test_size=test_size,
random_state=seed)
# fit model no training data
#Parâmetros: https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingClassifier.html
model = GradientBoostingClassifier(loss='exponential',
learning_rate=0.1,
random_state=0,
n_estimators=n_estimators)
model.fit(X_train, y_train)
# make predictions for test data
start_time = utils.get_time() # Tempo inicial
y_pred = model.predict(X_test)
predictions = [round(value) for value in y_pred]
diff_time = utils.get_time_diff(start_time) # Tempo final
print_resultados(model, y_test, predictions, diff_time, X_test, nome,
nomeFeature)
def Classifier_KNN(X, y, resp1):
nome = "KNN"
X_new, nomeFeature = selecao_feature(X, y, resp1)
# split data into train and test sets
seed = 100
test_size = 0.2
# separação treino - teste: 80 - 20
X_train, X_test, y_train, y_test = train_test_split(X_new,
y,
test_size=test_size,
random_state=seed)
# fit model no training data
model = KNeighborsClassifier(n_neighbors=3, metric='euclidean')
model.fit(X_train, y_train)
# make predictions for test data
start_time = utils.get_time() # Tempo inicial
y_pred = model.predict(X_test)
predictions = [round(value) for value in y_pred]
diff_time = utils.get_time_diff(start_time) # Tempo final
print_resultados(model, y_test, predictions, diff_time, X_test, nome,
nomeFeature)
def SVM(X, y, resp1):
nome = "SVM"
X_new, nomeFeature = selecao_feature(X, y, resp1)
# separação treino - teste: 80 - 20
X_train, X_test, y_train, y_test = train_test_split(X_new,
y,
test_size=0.2,
random_state=0)
# Treinamento e Predição
classifier = svm.SVC(kernel='rbf', C=5, gamma=0.0001, probability=True)
classifier.fit(X_train, y_train)
start_time = utils.get_time() # Tempo inicial
y_pred = classifier.predict(X_test)
diff_time = utils.get_time_diff(start_time) # Tempo final
print_resultados(classifier, y_test, y_pred, diff_time, X_test, nome,
nomeFeature)
def Random_Forest(X, y, resp1):
nome = "Ramdom_Forest"
X_new, nomeFeature = selecao_feature(X, y, resp1)
# separação treino - teste: 80 - 20
X_train, X_test, y_train, y_test = train_test_split(X_new,
y,
test_size=0.2,
random_state=0)
# Treinamento e Predição
# 20 árvores (com 100 árvores tem o mesmo resultado)
classifier = RandomForestClassifier(n_estimators=200, random_state=0)
classifier.fit(X_train, y_train)
# Predict
start_time = utils.get_time() # Tempo inicial
y_pred = classifier.predict(X_test)
diff_time = utils.get_time_diff(start_time) # Tempo final
print_resultados(classifier, y_test, y_pred, diff_time, X_test, nome,
nomeFeature)