# Hyperparameters tuning

# grid search for No.of epochs, batch size, optimizer, Learning Rate, Momentum,
# Network Weight Initialization, Dropout Regularization, Number of Neurons in the Hidden Layer

epochss = [10, 50, 100, 150]
batchess = [5, 10, 20, 40, 60, 80, 100]
learn_rss = [0.001, 0.01, 0.1, 0.2, 0.3]
momentumss = [0.0, 0.2, 0.4, 0.6, 0.8, 0.9]
dropout_rss = [0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9]
#weight_constraint = [1, 2, 3, 4, 5]
num_neuronss = [1, 5, 10, 15, 20, 25, 30]

optimizerss = ['RMSprop','Adagrad','SGD','Adadelta','Adam','Adamax','Nadam']
#activationss = ['softmax','softplus','softsign','relu','tanh','sigmoid','selu','hard_sigmoid','linear']
activationss = ['relu','selu','linear']

kernel_initss = ['glorot_uniform','glorot_normal','normal','uniform',
                      'lecun_uniform','zero','he_normal','he_uniform']

kernel_lambdass = [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.05, 0.1, 0.5]


param_grid = dict(optimizer=optimizers, epochs=epochs, batch_size=batches, init=inits)

grid = GridSearchCV(estimator=model, param_grid=param_grid)
grid_result = grid.fit(X, Y)