def training_data_generator(training_apps,
feats,
malwares,
path,
batch_size=64):
# training_apps = np.random.choice(train_test_apps, int(len(train_test_apps) * 0.66)) # 66% for training
gen_state = 0
while 1:
if gen_state + batch_size > len(training_apps):
apps = training_apps[gen_state:len(training_apps)]
y = []
for app in apps:
if app in malwares:
y.append(np.array([1, 0])) # malware
else:
y.append(np.array([0, 1])) # benign
X = [preprocess_app(app, feats, path) for app in apps]
gen_state = 0
else:
apps = training_apps[gen_state:gen_state + batch_size]
y = []
for app in apps:
if app in malwares:
y.append(np.array([1, 0])) # malware
else:
y.append(np.array([0, 1])) # benign
X = [preprocess_app(app, feats, path) for app in apps]
gen_state = gen_state + batch_size
yield np.array(X), np.array(y)
def testing_data(train_test_apps, feats, malwares, path):
testing_apps = np.random.choice(train_test_apps, int(len(train_test_apps) * 0.34)) # 34% for testing
xs = []
ys = []
for testing_app in testing_apps:
if testing_app in malwares:
ys.append(np.array([1, 0])) # malware
else:
ys.append(np.array([0, 1])) # benign
xs.append(preprocess_app(testing_app, feats, path))
xs = np.array(xs)
ys = np.array(ys)
np.save('testing_xs', xs)
np.save('testing_ys', ys)
return xs, ys