def test_SSelect():
tt = get_traintest()
x_train, y_train, x_test, y_test, path = next(tt)
x_train, y_train, x_test, y_test, path = next(tt)
SSelect_score = SSelect.SSelect(x_train, y_train, x_test)
SSelect_score_rank = SSelect.feature_ranking(SSelect_score)
num_fea = 100 # number of selected features
selected_fea = SSelect_score_rank[:num_fea]
clf = svm.LinearSVC()
clf.fit(x_train[:, selected_fea], y_train)
y_predict = clf.predict(x_test[:, selected_fea])
accuracy = accuracy_score(y_test, y_predict)
print('Accuracy : {0}'.format(accuracy))
def test_lsdf():
tt = get_traintest()
x_train, y_train, x_test, y_test, path = next(tt)
lsdf_score = lsdf.lsdf(x_train, y_train, x_test)
lsdf_score_rank = lsdf.feature_ranking(lsdf_score)
num_fea = 100 # number of selected features
idx = lsdf_score_rank[:num_fea]
accuracy = 0
run_num = 10
for i in range(run_num):
clf = svm.LinearSVC()
clf.fit(x_train[:, idx], y_train)
y_predict = clf.predict(x_test[:, idx])
accuracy += accuracy_score(y_test, y_predict)
print('Accuracy : {0}'.format(accuracy/run_num))
def cal_baseline():
num_folders = 3
output_path = './result/'
fn = 'baseline_accuracy_folders_{0}.txt'.format(num_folders)
# load data
gd = get_data()
for X, y, path in gd:
ss = cross_validation.StratifiedKFold(y,
n_folds=num_folders,
shuffle=True)
# perform evaluation on classification task
clf = svm.LinearSVC() # linear SVM
correct = 0
for train, test in ss:
# obtain the score of each feature on the training set
score = fisher_score.fisher_score(X[train], y[train])
# rank features in descending order according to score
idx = fisher_score.feature_ranking(score)
# train a classification model with the selected features on the training dataset
clf.fit(X[train], y[train])
# predict the class labels of samples2 data
y_predict = clf.predict(X[test])
# obtain the classification accuracy on the samples2 data
acc = accuracy_score(y[test], y_predict)
correct += acc
# output the average classification accuracy over all k folds
avg_accuracy = correct * 1.0 / num_folders
new_path = output_path + path.split('data')[-1].strip()
create_path(new_path)
with open(new_path + '/' + fn, 'w+') as f:
print(avg_accuracy, file=f)
print('{0} finish!'.format(__file__))
def main():
# load data
gd = get_data()
X,y, path = next(gd)
n_samples, n_features = X.shape # number of samples2 and number of features
# split data into several folds
num_folders = 3
ss = cross_validation.StratifiedKFold(y, n_folds=num_folders, shuffle=True)
# perform evaluation on classification task
num_fea = 100 # number of selected features
clf = svm.LinearSVC() # linear SVM
correct = 0
for train, test in ss:
# obtain the score of each feature on the training set
score = fisher_score.fisher_score(X[train], y[train])
# rank features in descending order according to score
idx = fisher_score.feature_ranking(score)
# obtain the dataset on the selected features
selected_features = X[:, idx[0:num_fea]]
# train a classification model with the selected features on the training dataset
clf.fit(selected_features[train], y[train])
# predict the class labels of samples2 data
y_predict = clf.predict(selected_features[test])
# obtain the classification accuracy on the samples2 data
acc = accuracy_score(y[test], y_predict)
correct += acc
# output the average classification accuracy over all 10 folds
print('Accuracy:', float(correct)/num_folders)