def adaboost(X, Y, X_test, Y_test, cv=10):
clf = AdaBoostClassifier(n_estimators=100, learning_rate=0.2)
return crossValidation(clf, X, Y)
from knn import * from DataAcesses import * from CrossValidation import crossValidation i = Database() dados = i.selectAll() crossValidation(dados, 2)
def logisticRegression(X, Y, X_test, Y_test, cv=10):
clf = LogisticRegression(penalty="l1", C=2.0)
return crossValidation(clf, X, Y)
def bagging(X, Y, X_test, Y_test):
clf = BaggingClassifier(n_estimators=100,
max_features=0.5,
bootstrap=False)
return crossValidation(clf, X, Y)
def deep(X, Y, X_test, Y_test, cv=10):
clf = MLPClassifier(hidden_layer_sizes=(40, 30, 20, 10, 5, 2, 1),
learning_rate="invscaling")
return crossValidation(clf, X, Y)
def perceptron(X, Y, X_test, Y_test, cv=10):
clf = Perceptron(penalty='l1', alpha=0.0001)
return crossValidation(clf, X, Y)
def neuralNet(X, Y, X_test, Y_test, cv=10):
clf = MLPClassifier(hidden_layer_sizes=(
30, 15), learning_rate="invscaling")
return crossValidation(clf, X, Y)
def decisionTree(X, Y, X_test, Y_test, cv=10):
clf = tree.DecisionTreeClassifier(
max_depth=20, min_samples_split=25, max_leaf_nodes=100)
return crossValidation(clf, X, Y)
def gradientBoost(X, Y, X_test, Y_test, cv=10):
clf = GradientBoostingClassifier(
learning_rate=0.1, n_estimators=200, max_depth=4)
return crossValidation(clf, X, Y)
def mnb(X, Y, X_test, Y_test, cv=10):
clf = MultinomialNB(alpha=1.0)
return crossValidation(clf, X, Y)
def svc(X, Y, X_test, Y_test, cv=10):
clf = SVC(kernel='linear', C=0.4, degree=2)
return crossValidation(clf, X, Y)
from knn import *
from DataAcesses import *
from CrossValidation import crossValidation
i = Database()
dados = i.selectAll()
for a in range(2, 120):
print(str(a) + ' : ' + str(crossValidation(dados, a)))
def knn(X, Y, X_test, Y_test, cv=10):
clf = KNeighborsClassifier(n_neighbors=5, leaf_size=30, p=2)
return crossValidation(clf, X, Y)
def randomForest(X, Y, X_test, Y_test, cv=10):
clf = RandomForestClassifier(n_estimators=150,
max_depth=None,
max_leaf_nodes=None)
return crossValidation(clf, X, Y)