def create_model(net, model_type, save_model):
net = regression(net,
optimizer='adam',
loss='categorical_crossentropy',
name='output')
if save_model:
return DNN(net, checkpoint_path=get_checkpoint_path(model_type))
else:
return DNN(net)
def create_model(net):
net = regression(net,
optimizer='adam',
loss='categorical_crossentropy',
name='output')
return DNN(net)
def createModel(self):
print("Creating model...")
imageSize = self.config.sequenceLength
nbClasses = self.config.nbClasses
net = input_data(shape=[None, imageSize, imageSize, 1], name='input')
net = conv_2d(net, imageSize/2, 2, activation='elu', weights_init="Xavier", name='conv1')
net = max_pool_2d(net, 2)
net = conv_2d(net, imageSize, 2, activation='elu', weights_init="Xavier", name='conv2')
net = max_pool_2d(net, 2)
net = conv_2d(net, imageSize*2, 2, activation='elu', weights_init="Xavier", name='conv3')
net = max_pool_2d(net, 2)
net = conv_2d(net, imageSize*4, 2, activation='elu', weights_init="Xavier", name='conv4')
net = max_pool_2d(net, 2)
net = fully_connected(net, imageSize*8, activation='elu')
net = dropout(net, 0.5)
net = fully_connected(net, nbClasses, activation='softmax')
net = regression(net, optimizer='rmsprop', loss='categorical_crossentropy')
self.model = DNN(net)
print("Model created!")
def train(self):
""" Start training """
# 1: build a list of image filenames
self.build_image_filenames_list()
# 2: use list information to init our numpy variables
self.init_np_variables()
# 3: Add images to our Tensorflow dataset
self.add_tf_dataset(self.list_cow_files, 0)
self.add_tf_dataset(self.list_noncow_files, 1)
# 4: Process TF dataset
self.process_tf_dataset()
# 5: Setup image preprocessing
self.setup_image_preprocessing()
# 6: Setup network structure
self.setup_nn_network()
# 7: Train our deep neural network
tf_model = DNN(self.tf_network, tensorboard_verbose=3,
checkpoint_path='model_cows.tfl.ckpt')
tf_model.fit(self.tf_x, self.tf_y, n_epoch=100, shuffle=True,
validation_set=(self.tf_x_test, self.tf_y_test),
show_metric=True, batch_size=96,
snapshot_epoch=True,
run_id='model_cows')
# 8: Save model
tf_model.save('model_cows.tflearn')
def __init__(self):
# initializebevideo stream
self.video_stream = cv2.VideoCapture(
VIDEO_PREDICTOR.camera_source)
self.face_detector = cv2.CascadeClassifier(
VIDEO_PREDICTOR.face_detection_classifier)
self.shape_predictor = None
self.shape_predictor = dlib.shape_predictor(
DATASET.shape_predictor_path)
model = None
with tf.Graph().as_default():
network = build_model(
use_landmarks=True,
use_hog_and_landmarks=True,
use_hog_sliding_window_and_landmarks=True)
model = DNN(network)
if os.path.isfile(
TRAINING.save_model_path_landmarks_hog_sw):
model.load(TRAINING.save_model_path_landmarks_hog_sw)
else:
print("Error: file '{}' not found".format(
TRAINING.save_model_path_landmarks_hog_sw))
self.model = model
self.last_predicted_time = 0
self.last_predicted_confidence = 0
self.last_predicted_emotion = ""
def _build_network(self):
self.logger.info('Started CNN structure construction')
network = input_data(shape=[None, self.height, self.width, 3],
dtype=float32)
network = conv_2d(network, 64, 5, strides=4, activation='relu')
network = max_pool_2d(network, 3, strides=2)
network = local_response_normalization(network)
network = conv_2d(network, 128, 4, activation='relu')
network = max_pool_2d(network, 3, strides=2)
network = local_response_normalization(network)
network = conv_2d(network, 256, 1, activation='relu')
network = max_pool_2d(network, 3, strides=2)
network = local_response_normalization(network)
network = fully_connected(network, 4096, activation='tanh')
network = dropout(network, 0.5)
network = fully_connected(network, 4096, activation='tanh')
network = dropout(network, 0.5)
network = fully_connected(network, 2, activation='softmax')
network = regression(network,
optimizer='momentum',
loss='categorical_crossentropy',
learning_rate=0.001)
self.cnn_ = DNN(network,
checkpoint_path='firenet',
max_checkpoints=1,
tensorboard_verbose=2)
self.logger.info('Finished CNN structure construction')
def _create_model():
reset_default_graph()
net = input_data([None, 5])
net = fully_connected(net, N_HIDDEN_UNITS, bias=True, activation='tanh')
net = fully_connected(net, 2, activation='softmax')
net = regression(net, optimizer='adam', learning_rate=0.001,
loss='categorical_crossentropy')
return DNN(net)
def load_model(self):
"""
This method loads a created model from the drive.
"""
model = None
with tf.Graph().as_default():
self.create_network()
model = DNN(self.network, tensorboard_verbose=0)
model.load("./classifier/models/classifier.tflearn")
self.model = model
def BotANN():
tf.reset_default_graph()
rete = input_data(shape=[None, len(temi)])
rete = fully_connected(rete, 8)
rete = fully_connected(rete, 8)
rete = fully_connected(rete, len(classi), activation='softmax')
rete = regression(rete)
model = DNN(rete, tensorboard_dir='logs')
return model
def train(
self,
X_train,
Y_train,
X_val,
Y_val
):
with tf.Graph().as_default():
print("Building Model...........")
network = build_CNN()
model = DNN(
network,
tensorboard_dir="path_to_logs",
tensorboard_verbose=0,
checkpoint_path="path_to_checkpoints",
max_checkpoints=1
)
if self.is_training:
# Training phase
print("start training...")
print(" - emotions = {}".format(7))
print(" - optimizer = '{}'".format(self.optimizer))
print(" - learning_rate = {}".format(0.016))
print(" - learning_rate_decay = {}".format(self.learning_rate_decay))
print(" - otimizer_param ({}) = {}".format(self.optimizer, self.optimizer_param))
print(" - Dropout = {}".format(self.dropout))
print(" - epochs = {}".format(self.epochs))
start_time = time.time()
model.fit(
{'input': X_train.reshape(-1, 48, 48, 1)},
{'output': Y_train},
validation_set=(
{'input': X_val.reshape(-1, 48, 48, 1)},
{'output': Y_val},
),
batch_size=128,
n_epoch=10,
show_metric=True,
snapshot_step=100
)
training_time = time.time() - start_time
print("training time = {0:.1f} sec".format(training_time))
print("saving model...")
model.save("saved_model.bin")
def load_model():
model = None
with tf.Graph().as_default():
print "loading pretrained model..."
network = build_model()
model = DNN(network)
if os.path.isfile(TRAINING.save_model_path):
model.load(TRAINING.save_model_path)
else:
print "Error: file '{}' not found".format(TRAINING.save_model_path)
return model
def create_modelTF():
network = input_data(shape=[None, 5, 1], name='input')
network = fully_connected(network, 25, activation='relu')
network = fully_connected(network, 1, activation='linear')
network = regression(network,
optimizer='adam',
learning_rate=1e-2,
loss='mean_square',
name='target')
# model = DNN(network, checkpoint_path='snake_nn.tfl', tensorboard_dir='log', max_checkpoints=1 )
model = DNN(network, tensorboard_dir='log')
return model
def setup(self):
ops.reset_default_graph()
net = input_data(shape=[None, len(self.training[0])])
net = fully_connected(net, 10)
net = fully_connected(net, 10)
net = fully_connected(net, len(self.output[0]), activation="softmax")
net = regression(net)
self.model = DNN(net)
if exists("src/models/model.tflearn.index"):
self.model.load("src/models/model.tflearn")
else:
self.model.fit(self.training, self.output, n_epoch=1000, batch_size=8, show_metric=True)
self.model.save("src/models/model.tflearn")
def __build_model(self):
"""
Composes a neural network architecture capable of predicting the level of a student given his/her grades.
:return: The neural network model.
"""
tf.compat.v1.reset_default_graph()
nn = input_data(shape=[None, 3])
nn = fully_connected(nn, 32)
nn = fully_connected(nn, 32)
nn = fully_connected(nn, 10, activation='softmax')
nn = regression(nn)
model = DNN(nn)
return model
def _create_model(self):
reset_default_graph()
net = input_data([None, SEQUENCE_LEN])
net = embedding(net,
input_dim=len(self._vocab.vocabulary_),
output_dim=WORD_FEATURE_DIM)
net = lstm(net, DOC_FEATURE_DIM, dropout=0.8)
net = fully_connected(net, 2, activation='softmax')
net = regression(net,
optimizer='adam',
learning_rate=0.001,
loss='categorical_crossentropy')
return DNN(net)
def build(self, checkpoint_path, max_checkpoints, tensorboard_verbose):
"""
Args:
checkpoint_path (str):
max_checkpoints (int):
tensorboard_verbose (int):
Returns
tflearn.DNN: A tflearn DNN object that can be used as an estimator.
"""
return DNN(self.network, checkpoint_path=checkpoint_path,
max_checkpoints=max_checkpoints,
tensorboard_verbose=tensorboard_verbose)
def load_model():
model = None
with tf.Graph().as_default():
print("loading pretrained model...")
network = build_model(use_landmarks=True,
use_hog_and_landmarks=True,
use_hog_sliding_window_and_landmarks=True)
model = DNN(network)
if os.path.isfile(TRAINING.save_model_path_landmarks_hog_sw):
model.load(TRAINING.save_model_path_landmarks_hog_sw)
else:
print("Error: file '{}' not found".format(
TRAINING.save_model_path_landmarks_hog_sw))
return model
def createModel(self):
print("Creating model...")
sequenceLength = self.config.sequenceLength
nbClasses = self.config.nbClasses
nFeatures = self.config.nFeatures
net = input_data(shape=[None, sequenceLength, nFeatures])
net = lstm(net, n_units=nFeatures, dropout=0.8)
net = fully_connected(net, nbClasses, activation='softmax')
net = regression(net)
self.model = DNN(net)
print("Model created!")
def model(size=32, color_channels=3):
convnet = input_data((None, size, size, color_channels))
convnet = conv_2d(convnet, 16, 5, activation='relu')
convnet = conv_2d(convnet, 16, 5, activation='relu')
convnet = max_pool_2d(convnet, 2)
convnet = conv_2d(convnet, 20, 3, activation='relu')
convnet = max_pool_2d(convnet, 2)
convnet = conv_2d(convnet, 20, 3, activation='relu')
convnet = max_pool_2d(convnet, 2)
convnet = fully_connected(convnet, 1024, 'relu')
convnet = dropout(convnet, .9)
convnet = fully_connected(convnet, 10, 'softmax')
convnet = regression(convnet)
return DNN(convnet, tensorboard_verbose=3)
def tflearn_cifar():
"""
图像分类
:return:
"""
(X_train, Y_train), (X_test, Y_test) = cifar10.load_data()
X_train, Y_train = shuffle(X_train, Y_train)
Y_train = to_categorical(Y_train, nb_classes=10)
Y_test = to_categorical(Y_test, nb_classes=10)
# 对数据集进行零中心化（即对整个数据集计算平均值），同时进行 STD 标准化（即对整个数据集计算标准差）
img_prep = ImagePreprocessing()
img_prep.add_featurewise_zero_center()
img_prep.add_featurewise_stdnorm()
# 通过随机左右翻转和随机旋转来增强数据集
img_aug = ImageAugmentation()
img_aug.add_random_flip_leftright()
img_aug.add_random_rotation(max_angle=25.)
# 定义模型
network = input_data(shape=(None, 32, 32, 3),
data_preprocessing=img_prep,
data_augmentation=img_aug)
network = conv_2d(network, 32, 3, activation="relu")
network = max_pool_2d(network, 2)
network = conv_2d(network, 64, 3, activation="relu")
network = conv_2d(network, 64, 3, activation="relu")
network = max_pool_2d(network, 2)
network = fully_connected(network, 512, activation="relu")
network = dropout(network, 0.5)
network = fully_connected(network, 10, activation="softmax")
network = regression(network,
optimizer="adam",
loss="categorical_crossentropy",
learning_rate=0.001)
# 训练模型
model = DNN(network, tensorboard_verbose=0)
model.fit(X_train,
Y_train,
n_epoch=50,
shuffle=True,
validation_set=(X_test, Y_test),
show_metric=True,
batch_size=96,
run_id="cifar10_cnn")
def create_model(self):
"""
Creates DNN model that is based on built algorithm.
Needed algorithm is builded with self.build_algorithm call.
"""
self.log_named("model creation started")
if self.algorithm is not None:
self.model = DNN(self.algorithm,
checkpoint_path=self.checkpoints_dir_path,
max_checkpoints=1,
tensorboard_verbose=3,
tensorboard_dir=self.learn_logs_dir_path)
self.log_named("model creation finished")
else:
self.log_named_warning(
"model was not created, because algorithm is None!")
def use_tflearn(x_train, y_train, x_test, y_test):
net = input_data(shape=[None, x_train.shape[1]], name='input')
net = fully_connected(net, 24, activation='sigmoid', bias_init='normal')
net = fully_connected(net, 16, activation='sigmoid', bias_init='normal')
net = fully_connected(net, 12, activation='sigmoid', bias_init='normal')
net = fully_connected(net, 8, activation='sigmoid', bias_init='normal')
net = regression(net)
model = DNN(net,
tensorboard_dir=TENSORBOARD_DIR.as_posix(),
tensorboard_verbose=3,
best_checkpoint_path=CHECKPOINT_PATH.as_posix())
model.fit(x_train, y_train,
validation_set=(x_test, y_test),
n_epoch=100,
batch_size=10,
show_metric=True,
run_id='DNN-4f')
model.save(MODEL_FILE.as_posix())
return model
def createModel(self):
print("Creating model...")
nFeatures = self.config.nFeatures
nbClasses = self.config.nbClasses
#preprocess = DataPreprocessing()
#preprocess.add_featurewise_zero_center()
#preprocess.add_featurewise_stdnorm()
net = input_data(shape=[None, nFeatures])
net = fully_connected(net, 32, activation='elu', weights_init="Xavier")
net = fully_connected(net, 64, activation='elu', weights_init="Xavier")
net = fully_connected(net,
128,
activation='elu',
weights_init="Xavier")
net = fully_connected(net, nbClasses, activation='softmax')
net = regression(net)
self.model = DNN(net, tensorboard_verbose=3)
print("Model created!")
def load_model():
model = None
with tf.Graph().as_default():
print("loading pretrained model...")
data, validation, test = load_data(validation=True, test=True)
network = build_model()
model = DNN(network)
if os.path.isfile(TRAINING.save_model_path):
model.load(TRAINING.save_model_path)
else:
print("Error: file '{}' not found".format(TRAINING.save_model_path))
print("--")
print("Validation samples: {}".format(len(validation['Y'])))
print("Test samples: {}".format(len(test['Y'])))
print("--")
print("evaluating...")
start_time = time.time()
validation_accuracy = evaluate(model, validation['X'], validation['X2'], validation['Y'])
print(" - validation accuracy = {0:.1f}".format(validation_accuracy * 100))
test_accuracy = evaluate(model, test['X'], test['X2'], test['Y'])
print(" - test accuracy = {0:.1f}".format(test_accuracy * 100))
print(" - evalution time = {0:.1f} sec".format(time.time() - start_time))
return model
# -*- coding: utf-8 -*-
"""
Source : https://towardsdatascience.com/tflearn-soving-xor-with-a-2x2x1-feed-forward-neural-network-6c07d88689ed
"""
from tflearn import DNN
from tflearn.layers.core import input_data, dropout, fully_connected
from tflearn.layers.estimator import regression
#Training examples
X = [[0,0], [0,1], [1,0], [1,1]]
Y = [[0], [1], [1], [0]]
input_layer = input_data(shape=[None, 2]) #input layer of size 2
hidden_layer = fully_connected(input_layer , 2, activation='tanh') #hidden layer of size 2
output_layer = fully_connected(hidden_layer, 1, activation='tanh') #output layer of size 1
#use Stohastic Gradient Descent and Binary Crossentropy as loss function
regression = regression(output_layer , optimizer='sgd', loss='binary_crossentropy', learning_rate=5)
model = DNN(regression)
#fit the model
model.fit(X, Y, n_epoch=5000, show_metric=True);
#predict all examples
print ('Expected: ', [i[0] > 0 for i in Y])
print ('Predicted: ', [i[0] > 0 for i in model.predict(X)])
model.get_weights(hidden_layer.W)
model.get_weights(output_layer.W)
model.save("tflearn-xor")
fully_connected(net[i], 1, activation='sigmoid') for i in range(len(net))
]
net = merge(net, mode='concat')
print("After RNN : ", net.get_shape().as_list())
print("After Dropout : ", net.get_shape().as_list())
net = regression(net,
optimizer='adam',
loss='binary_crossentropy',
learning_rate=0.005)
print("After regression : ", net.get_shape().as_list())
testX = trainX[int(0.3 * len(trainY)):]
testY = trainY[int(0.3 * len(trainY)):]
# Training
model = DNN(net, clip_gradients=0., tensorboard_verbose=2)
embeddingWeights = get_layer_variables_by_name('EmbeddingLayer')[0]
# Assign your own weights (for example, a numpy array [input_dim, output_dim])
model.set_weights(embeddingWeights, embeddings)
model.fit(trainX,
trainY,
n_epoch=3,
validation_set=0.1,
show_metric=True,
batch_size=32,
shuffle=True)
#print( model.evaluate(testX, testY) )
predictions = model.predict(testX)
predictions = prob2Onehot(predictions)
#print("Predictions : ", list(predictions[10]))
def train(optimizer=HYPERPARAMS.optimizer,
optimizer_param=HYPERPARAMS.optimizer_param,
learning_rate=HYPERPARAMS.learning_rate,
keep_prob=HYPERPARAMS.keep_prob,
learning_rate_decay=HYPERPARAMS.learning_rate_decay,
decay_step=HYPERPARAMS.decay_step,
train_model=True):
print "loading dataset " + DATASET.name + "..."
if train_model:
data, validation = load_data(validation=True)
else:
data, validation, test = load_data(validation=True, test=True)
with tf.Graph().as_default():
print "building model..."
network = build_model(optimizer, optimizer_param, learning_rate,
keep_prob, learning_rate_decay, decay_step)
model = DNN(network,
tensorboard_dir=TRAINING.logs_dir,
tensorboard_verbose=0,
checkpoint_path=TRAINING.checkpoint_dir,
max_checkpoints=TRAINING.max_checkpoints)
#tflearn.config.init_graph(seed=None, log_device=False, num_cores=6)
if train_model:
# Training phase
print "start training..."
print " - emotions = {}".format(NETWORK.output_size)
print " - optimizer = '{}'".format(optimizer)
print " - learning_rate = {}".format(learning_rate)
print " - learning_rate_decay = {}".format(learning_rate_decay)
print " - otimizer_param ({}) = {}".format(
'beta1' if optimizer == 'adam' else 'momentum',
optimizer_param)
print " - keep_prob = {}".format(keep_prob)
print " - epochs = {}".format(TRAINING.epochs)
print " - use landmarks = {}".format(NETWORK.use_landmarks)
print " - use hog + landmarks = {}".format(
NETWORK.use_hog_and_landmarks)
print " - use hog sliding window + landmarks = {}".format(
NETWORK.use_hog_sliding_window_and_landmarks)
print " - use batchnorm after conv = {}".format(
NETWORK.use_batchnorm_after_conv_layers)
print " - use batchnorm after fc = {}".format(
NETWORK.use_batchnorm_after_fully_connected_layers)
start_time = time.time()
if NETWORK.use_landmarks:
model.fit([data['X'], data['X2']],
data['Y'],
validation_set=([validation['X'],
validation['X2']], validation['Y']),
snapshot_step=TRAINING.snapshot_step,
show_metric=TRAINING.vizualize,
batch_size=TRAINING.batch_size,
n_epoch=TRAINING.epochs)
else:
model.fit(data['X'],
data['Y'],
validation_set=(validation['X'], validation['Y']),
snapshot_step=TRAINING.snapshot_step,
show_metric=TRAINING.vizualize,
batch_size=TRAINING.batch_size,
n_epoch=TRAINING.epochs)
validation['X2'] = None
training_time = time.time() - start_time
print "training time = {0:.1f} sec".format(training_time)
if TRAINING.save_model:
print "saving model..."
model.save(TRAINING.save_model_path)
if not(os.path.isfile(TRAINING.save_model_path)) and \
os.path.isfile(TRAINING.save_model_path + ".meta"):
os.rename(TRAINING.save_model_path + ".meta",
TRAINING.save_model_path)
print "evaluating..."
validation_accuracy = evaluate(model, validation['X'],
validation['X2'], validation['Y'])
print " - validation accuracy = {0:.1f}".format(
validation_accuracy * 100)
return validation_accuracy
else:
# Testing phase : load saved model and evaluate on test dataset
print "start evaluation..."
print "loading pretrained model..."
if os.path.isfile(TRAINING.save_model_path):
model.load(TRAINING.save_model_path)
else:
print "Error: file '{}' not found".format(
TRAINING.save_model_path)
exit()
if not NETWORK.use_landmarks:
validation['X2'] = None
test['X2'] = None
print "--"
print "Validation samples: {}".format(len(validation['Y']))
print "Test samples: {}".format(len(test['Y']))
print "--"
print "evaluating..."
start_time = time.time()
validation_accuracy = evaluate(model, validation['X'],
validation['X2'], validation['Y'])
print " - validation accuracy = {0:.1f}".format(
validation_accuracy * 100)
test_accuracy = evaluate(model, test['X'], test['X2'], test['Y'])
print " - test accuracy = {0:.1f}".format(test_accuracy * 100)
print " - evalution time = {0:.1f} sec".format(time.time() -
start_time)
return test_accuracy
network = input_data(shape=[None, 48, 48, 1], name='input')
network = conv_2d(network, nb_filter=32, filter_size=[7, 7], activation='relu')
network = max_pool_2d(network, kernel_size=2)
network = conv_2d(network, nb_filter=64, filter_size=[7, 7], activation='relu')
network = max_pool_2d(network, kernel_size=2)
network = fully_connected(network, n_units=512, activation='relu')
network = dropout(network, 0.5)
network = fully_connected(network, n_units=7, activation='softmax')
network = regression(network,
optimizer='adam',
learning_rate=0.01,
loss='categorical_crossentropy',
name='targets')
# abstract the network into a file
model = DNN(network)
model.load("model/model.tfl")
def predict(input_image):
prediction = model.predict(input_image)
print(prediction)
predicted_class = np.argmax(prediction)
print(predicted_class)
meaning_list = [
'Angry', 'Disgust', 'Fear', 'Happy', 'Sad', 'Surprise', 'Neutral'
]
print(meaning_list[predicted_class])
def tflearn_OneClass_NN_linear(data_train, data_test, labels_train):
X = data_train
Y = labels_train
D = X.shape[1]
No_of_inputNodes = X.shape[1]
# Clear all the graph variables created in previous run and start fresh
tf.reset_default_graph()
# Define the network
input_layer = input_data(shape=[None,
No_of_inputNodes]) # input layer of size
np.random.seed(42)
theta0 = np.random.normal(0, 1, K + K * D + 1) * 0.0001
#theta0 = np.random.normal(0, 1, K + K*D + 1) # For linear
hidden_layer = fully_connected(
input_layer,
4,
bias=False,
activation='linear',
name="hiddenLayer_Weights",
weights_init="normal") # hidden layer of size 2
output_layer = fully_connected(
hidden_layer,
1,
bias=False,
activation='linear',
name="outputLayer_Weights",
weights_init="normal") # output layer of size 1
# Initialize rho
value = 0.01
init = tf.constant_initializer(value)
rho = va.variable(name='rho', dtype=tf.float32, shape=[], initializer=init)
rcomputed = []
auc = []
sess = tf.Session()
sess.run(tf.initialize_all_variables())
# print sess.run(tflearn.get_training_mode()) #False
tflearn.is_training(True, session=sess)
print sess.run(tflearn.get_training_mode()) #now True
temp = theta0[-1]
oneClassNN_Net = oneClassNN(output_layer,
v,
rho,
hidden_layer,
output_layer,
optimizer='sgd',
loss='OneClassNN_Loss',
learning_rate=1)
model = DNN(oneClassNN_Net, tensorboard_verbose=3)
model.set_weights(output_layer.W, theta0[0:K][:, np.newaxis])
model.set_weights(hidden_layer.W, np.reshape(theta0[K:K + K * D], (D, K)))
iterStep = 0
while (iterStep < 100):
print "Running Iteration :", iterStep
# Call the cost function
y_pred = model.predict(data_train) # Apply some ops
tflearn.is_training(False, session=sess)
y_pred_test = model.predict(data_test) # Apply some ops
tflearn.is_training(True, session=sess)
value = np.percentile(y_pred, v * 100)
tflearn.variables.set_value(rho, value, session=sess)
rStar = rho
model.fit(X, Y, n_epoch=2, show_metric=True, batch_size=100)
iterStep = iterStep + 1
rcomputed.append(rho)
temp = tflearn.variables.get_value(rho, session=sess)
# print "Rho",temp
# print "y_pred",y_pred
# print "y_predTest", y_pred_test
# g = lambda x: x
g = lambda x: 1 / (1 + tf.exp(-x))
def nnScore(X, w, V, g):
return tf.matmul(g((tf.matmul(X, w))), V)
# Format the datatype to suite the computation of nnscore
X = X.astype(np.float32)
X_test = data_test
X_test = X_test.astype(np.float32)
# assign the learnt weights
# wStar = hidden_layer.W
# VStar = output_layer.W
# Get weights values of fc2
wStar = model.get_weights(hidden_layer.W)
VStar = model.get_weights(output_layer.W)
# print "Hideen",wStar
# print VStar
train = nnScore(X, wStar, VStar, g)
test = nnScore(X_test, wStar, VStar, g)
# Access the value inside the train and test for plotting
# Create a new session and run the example
# sess = tf.Session()
# sess.run(tf.initialize_all_variables())
arrayTrain = train.eval(session=sess)
arrayTest = test.eval(session=sess)
# print "Train Array:",arrayTrain
# print "Test Array:",arrayTest
# plt.hist(arrayTrain-temp, bins = 25,label='Normal');
# plt.hist(arrayTest-temp, bins = 25, label='Anomalies');
# plt.legend(loc='upper right')
# plt.title('r = %1.6f- Sigmoid Activation ' % temp)
# plt.show()
pos_decisionScore = arrayTrain - temp
neg_decisionScore = arrayTest - temp
return [pos_decisionScore, neg_decisionScore]
def neural_net_trainer(net, learning_rate=0.001, best_checkpoint_path='model_best'):
net = regression(net, optimizer='adam', learning_rate= learning_rate, loss='categorical_crossentropy', shuffle_batches=True)
return DNN(network=net, clip_gradients=0., tensorboard_verbose=3, best_checkpoint_path=best_checkpoint_path, best_val_accuracy=0.999)