def train_predict(clf, X_train, y_train, X_test):
preds_br = np.zeros((X_test.shape[0], 9))
for i in range(0, 9):
clf.fit(X_train, y_train[:, i])
preds_br[:, i] = clf.predict_proba(X_test)[:, 1]
nn_preds = np.array([])
n_iter = 20
for i in range(n_iter):
nn_clf.fit(X_train, y_train)
s_preds = nn_clf.predict_proba(X_test)
nn_preds = nn_preds + s_preds if nn_preds.size else s_preds
nn_preds = (nn_preds / n_iter)
n_chains = 20
preds_cc = np.zeros((X_test.shape[0], 9))
for i in range(n_chains):
cc = ClassifierChain(clf)
cc.fit(X_train, y_train)
preds_cc = preds_cc + cc.predict(X_test)
preds_cc = preds_cc / n_chains
preds_br = (preds_br + 3*nn_preds + 2*preds_cc) / 6
return preds_br
def train_predict(clf, X_train, y_train, X_test):
preds_br = np.zeros((X_test.shape[0], 9))
for i in range(0, 9):
clf.fit(X_train, y_train[:, i])
preds_br[:, i] = clf.predict_proba(X_test)[:, 1]
nn_preds = np.array([])
n_iter = 20
for i in range(n_iter):
nn_clf.fit(X_train, y_train)
s_preds = nn_clf.predict_proba(X_test)
nn_preds = nn_preds + s_preds if nn_preds.size else s_preds
nn_preds = (nn_preds / n_iter)
n_chains = 20
preds_cc = np.zeros((X_test.shape[0], 9))
for i in range(n_chains):
cc = ClassifierChain(clf)
cc.fit(X_train, y_train)
preds_cc = preds_cc + cc.predict(X_test)
preds_cc = preds_cc / n_chains
preds_br = (preds_br + 3 * nn_preds + 2 * preds_cc) / 6
return preds_br
def train_predict(feature, clf, X_train, y_train, X_test):
preds_br = np.zeros((X_test.shape[0], 9))
for i in range(0, 9):
clf.fit(X_train, y_train[:, i])
preds_br[:, i] = clf.predict_proba(X_test)[:, 1]
np.save('test/' + feature + '_br', preds_br)
nn_preds = np.array([])
n_iter = 40
for i in range(n_iter):
nn_clf.fit(X_train, y_train)
s_preds = nn_clf.predict_proba(X_test)
nn_preds = nn_preds + s_preds if nn_preds.size else s_preds
nn_preds = (nn_preds / n_iter)
np.save('test/' + feature + '_nn', nn_preds)
n_chains = 40
preds_cc = np.zeros((X_test.shape[0], 9))
for i in range(n_chains):
cc = ClassifierChain(clf)
cc.fit(X_train, y_train)
preds_cc = preds_cc + cc.predict(X_test)
preds_cc = preds_cc / n_chains
np.save('test/' + feature + '_cc', preds_cc)
preds_br = (preds_br + 3 * nn_preds + 2 * preds_cc) / 6
return preds_br
def train_predict(fold, feature, clf, X_train, y_train, X_test, y_test):
preds_br = np.zeros((X_test.shape[0], 9))
for i in range(0, 9):
clf.fit(X_train, y_train[:, i])
preds_br[:, i] = clf.predict_proba(X_test)[:, 1]
np.save('val3/' + str(fold) + '_' + feature + '_br', preds_br)
nn_preds = np.array([])
n_iter = 10
for i in range(n_iter):
nn_clf.fit(X_train, y_train)
s_preds = nn_clf.predict_proba(X_test)
nn_preds = nn_preds + s_preds if nn_preds.size else s_preds
nn_preds = (nn_preds / n_iter)
np.save('val3/' + str(fold) + '_' + feature + '_nn', nn_preds)
n_chains = 10
preds_cc = np.zeros((X_test.shape[0], 9))
for i in range(n_chains):
cc = ClassifierChain(clf)
cc.fit(X_train, y_train)
preds_cc = preds_cc + cc.predict(X_test)
preds_cc = preds_cc / n_chains
np.save('val3/' + str(fold) + '_' + feature + '_cc', preds_cc)
np.save('val3/' + str(fold) + '_' + feature + '_test', y_test)
preds_br = (preds_br + 3*nn_preds + 2*preds_cc) / 6
return preds_br
def do_cross_validation(X, Y):
X, Y = shuffle(X, Y)
folds = 10
kf = KFold(n_splits=folds, shuffle=True)
cc = ClassifierChain()
zero_one_score = 0
hamming_score = 0
accuracy = 0
for train, test in kf.split(X):
X_train, y_train = X[train], Y[train]
X_test, y_test = X[test], Y[test]
cc.fit(X_train, y_train)
zero_one_score += cc.zero_one_loss_score(X_test, y_test)
hamming_score += cc.hamming_loss_score(X_test, y_test)
accuracy += cc.accuracy_score(X_test, y_test)
zero_one_score /= folds
hamming_score /= folds
accuracy /= folds
#print "0/1 Loss:", zero_one_score
#print "Hamming Loss:", hamming_score
#print "Accuracy:", accuracy
#print
return [zero_one_score, hamming_score, accuracy]