### Answer 3
from keras.models import Sequential
from keras.layers import Dense
from keras.wrappers.scikit_learn import KerasClassifier
from keras.wrappers.scikit_learn import KerasRegressor

from sklearn.model_selection import GridSearchCV
import numpy

tr_x = np.stack(top_train)
tr_y = np.stack(bot_train)

def create_model():
    model = Sequential()

    model.add(Dense(units = 100,  activation = 'relu', kernel_initializer = 'he_normal', input_dim = 1537))

    model.add(Dense(units = 1536, activation = 'linear', kernel_initializer = 'he_normal'))
    # Describe the loss and how it is optimized
    # TODO: May need to adjust learning rate parameter used by adam in cv
    model.compile(loss = 'mean_squared_error', optimizer = 'adam', metrics = ['accuracy'])
    return model

model = KerasRegressor(build_fn=create_model, verbose=0)
#batch_size = [5, 10, 15, 20, 40, 60, 80, 100]
#epochs = [10, 50, 100, 150]
batch_size = [5, 10]
epochs = [10, 20]
param_grid = dict(batch_size=batch_size, epochs=epochs)

grid = GridSearchCV(estimator=model, param_grid=param_grid, scoring='neg_mean_squared_error', cv=2, n_jobs=2)
grid_result = grid.fit(tr_x, tr_y)

# summarize results
print("Best: %f using %s" % (grid_result.best_score_, grid_result.best_params_))
means = grid_result.cv_results_['mean_test_score']
stds = grid_result.cv_results_['std_test_score']
params = grid_result.cv_results_['params']
for mean, stdev, param in zip(means, stds, params):
    print("%f (%f) with: %r" % (mean, stdev, param))