#First layer
L1 = LayerNN(size_in=inputSize,
size_out=numberOfFeatures,
sparsity=0.1,
beta=3,
weightDecay=3e-3,
activation=FunctionModel.Sigmoid)
#Second layer
L2 = LayerNN(size_in=numberOfFeatures,
size_out=inputSize,
weightDecay=3e-3,
activation=FunctionModel.Sigmoid)
#Compile all together
AE = TheanoNNclass(options, (L1, L2))
#Compile train and predict functions
AE.trainCompile()
AE.predictCompile()
#Normalise CV data from 0..255 to 0..1
#X_CV = DATA_CV[:, 1:].T / 255.0
#Empty list to collect CV errors
CV_error = []
#Let's iterate!
for i in xrange(iterations):
#Get miniBatch of defined size from whole DATA
#Let's try to train NN to see, how it solves such task
#NN part
#Common options for whole NN
options = OptionsStore(learnStep=0.05,
minibatch_size=dataSize,
CV_size=dataSize)
#Layer architecture
#We will use only one layer with 2 neurons on input and 1 on output
L1 = LayerNN(size_in=dataFeatures,
size_out=1,
activation=FunctionModel.Linear)
#Compile NN
NN = TheanoNNclass(options, (L1, ))
#Compile NN train
NN.trainCompile()
#Compile NN oredict
NN.predictCompile()
#Number of iterations (cycles of training)
iterations = 1000
#Set step to draw
drawEveryStep = 100
draw = False
#CV error accumulator (for network estimation)
trainData.X = DataMutate.deNormalizer(trainData.X, afterzero=8)
print trainData.X.shape, trainData.Y.shape
L1 = LayerNN(size_in=trainData.input,
size_out=numberOfFeatures,
sparsity=0.05,
beta=7,
weightDecay=3e-7,
activation=FunctionModel.Sigmoid)
L2 = LayerNN(size_in=numberOfFeatures,
size_out=trainData.input,
weightDecay=3e-7,
activation=FunctionModel.Sigmoid)
AE = TheanoNNclass(options, (L1, L2))
AE.trainCompile()
for i in xrange(iterations):
X, index = trainData.miniBatch(batchSize)
AE.trainCalc(X, X, iteration=1, debug=True)
print i
if os.path.exists(datasetFolder + "out/"):
shutil.rmtree(datasetFolder + "out/")
os.makedirs(datasetFolder + "out/")
os.makedirs(datasetFolder + "out/w/")
AE.modelSaver(datasetFolder + "out/" + "/AE_NEW_VERSION_MNIST.txt")
#First layer
L1 = LayerNN(size_in=inputSize,
size_out=numberOfFeatures,
sparsity=0.1,
beta=3,
weightDecay=3e-3,
activation=FunctionModel.Sigmoid)
#Second layer
L2 = LayerNN(size_in=numberOfFeatures,
size_out=inputSize,
weightDecay=3e-3,
activation=FunctionModel.Sigmoid)
#Compile all together
AE = TheanoNNclass(options, (L1, L2))
#Compile train and predict functions
AE.trainCompile()
AE.predictCompile()
#Normalise CV data from 0..255 to 0..1
X_CV = DATA_CV[:, 1:].T / 255.0
#Empty list to collect CV errors
CV_error = []
#Let's iterate!
for i in xrange(iterations):
#Get miniBatch of defined size from whole DATA
# First layer
L1 = LayerNN(
size_in=inputSize,
size_out=numberOfFeatures,
sparsity=0.1,
beta=3,
weightDecay=3e-3,
activation=FunctionModel.Sigmoid,
)
# Second layer
L2 = LayerNN(size_in=numberOfFeatures, size_out=inputSize, weightDecay=3e-3, activation=FunctionModel.Sigmoid)
# Compile all together
AE = TheanoNNclass(options, (L1, L2))
# Compile train and predict functions
AE.trainCompile()
AE.predictCompile()
# Normalise CV data from 0..255 to 0..1
X_CV = DATA_CV[:, 1:].T / 255.0
# Empty list to collect CV errors
CV_error = []
# Let's iterate!
for i in xrange(iterations):
# Get miniBatch of defined size from whole DATA
