minor fixes. update helper, autoencoder and main.py

[heykulthe]

Problem 1:

preprocessing.OneHotEncoder() function throws error message that the argument categorical_features is unexpected

Fix 1:

# Ensure categorical columns are properly encoded using OneHotEncoder
one_hot_encoder = preprocessing.OneHotEncoder(categories='auto', sparse=False, handle_unknown='ignore')
one_hot_encoded_cols = one_hot_encoder.fit_transform(dataset[:, categorical_cols])

Problem 2:

np.bool is depracated. Throws an error when running main.py

Fix 2:

Replace np.bool with bool

Problem 3:

When running with default model: autoencoder
accuracy function from autoencoder.py throws error:

temp = K.ones(K.shape(mse)) seems to be causing an out-of-scope error.

Fix 3:

Replace temp = K.ones(K.shape(mse)) with temp = K.ones_like(mse)