def keep_best_loss(X_train,y_train,X_valid,y_valid):
PROB,LOSS,RF = [],[],[]
for i in range(0,6):
seed=randrange(1,10000)
random.seed(seed)
print("Seed: %s" % seed)
rf = build_rf(X_train,y_train)
prob = rf.predict_proba(X_valid)
loss = of.logloss_mc(y_valid,prob)
RF.append(rf)
PROB.append(prob)
LOSS.append(loss)
print("Model %s Loss: %0.4f" % (i,loss))
sum=0
for p in PROB:
sum+=p
avg_prob = sum/len(PROB)
avg_loss = of.logloss_mc(y_valid,avg_prob)
print("Avg Loss of 5 models: %0.4f" % avg_loss)
return RF,PROB
def build_XB1(X_train,y_train,X_cal,y_cal,X_test):
cal_prob,test_prob = 0,0
for i in range(5):
print("--Building and Training model %s" % i)
seed = randrange(1,10000)
model = XGBoostClassifier(n_iter=500,
max_features=0.3,
max_depth=7,
min_child_weight=10,
gamma=0.0093,random_state=seed,
learning_rate=0.2,
l2_weight=0.1,
max_samples=0.9
)
model = CalibratedClassifierCV(base_estimator=model,method='isotonic',cv=5).fit(X_train,y_train)
print("Model %s training complete." % i)
cal_prob += model.predict_proba(X_cal)
test_prob += model.predict_proba(X_test)
cal_prob = cal_prob/5.
test_prob = test_prob/5.
print("Average Model Loss: %0.4f" % of.logloss_mc(y_cal,cal_prob))
return(cal_prob,test_prob)
def build_NN(X_train,y_train_nn,X_cal,y_cal,X_test):
cal_prob,test_prob = 0,0
nb_classes = y_train_nn.shape[1]
dims = X_train.shape[1]
for i in range(5):
print("--Building and Training model %s" % i)
seed = randrange(1,10000)
np.random.seed(seed)
model = Sequential()
model.add(Dense(dims, 512, init='glorot_uniform'))
model.add(PReLU((512,)))
model.add(BatchNormalization((512,)))
model.add(Dropout(0.5))
model.add(Dense(512, 512, init='glorot_uniform'))
model.add(PReLU((512,)))
model.add(BatchNormalization((512,)))
model.add(Dropout(0.5))
model.add(Dense(512, 512, init='glorot_uniform'))
model.add(PReLU((512,)))
model.add(BatchNormalization((512,)))
model.add(Dropout(0.5))
model.add(Dense(512, nb_classes, init='glorot_uniform'))
model.add(Activation('softmax'))
model.compile(loss='categorical_crossentropy', optimizer="adam")
model.fit(X_train, y_train_nn, nb_epoch=20, batch_size=16, validation_split=0.15)
cal_prob += model.predict_proba(X_cal)
test_prob += model.predict_proba(X_test)
cal_prob,test_prob = cal_prob/10.,test_prob/10.
print("Avg NN Model Loss: %0.4f" % of.logloss_mc(y_cal,cal_prob))
return(cal_prob,test_prob)