def main():
vae = VAE(input_dim, latent_dim)
input_x = tflearn.input_data(shape=(None, input_dim), name='input_x')
optimizer = tflearn.optimizers.Adam().get_tensor()
trainer = vae.return_trainer(input_x, optimizer, batch_size)
trainer.fit(feed_dicts={input_x: trainX},
val_feed_dicts={input_x: testX},
n_epoch=n_epoch,
shuffle_all=True,
run_id='VAE')
def main():
global trainX, trainY, testX, testY
vae = VAE(input_dim, latent_dim)
input_x = tflearn.input_data(shape=(None, input_dim), name='input_x')
optimizer = tflearn.optimizers.Adam().get_tensor()
trainer = vae.return_trainer(input_x, optimizer, batch_size)
trainer.restore(vae.get_checkpoint())
# calculate mu and logvar for trainX and testX
evaluator = vae.return_evaluator(trainer)
train_mu_logvar = evaluator.predict({input_x: trainX})
test_mu_logvar = evaluator.predict({input_x: testX})
# classification
classifier = SupportVectorClassifier()
classifier.fit(train_mu_logvar, trainY)
# evaluate
classifier.score(test_mu_logvar, testY)
