Hi I would like to propose a better implementation for 'test_indices':

We can remove the unneeded np.array casting:

Cleaner/New:
test_indices = list(set(range(len(texts))) - set(train_indices))

Old:
test_indices = np.array(list(set(range(len(texts))) - set(train_indices)))

Suggest that adding classification to LSTM and regressor to RandomForest on these two most popular models.

RandomForest is an awesome classification tool but it lacks of sequential data maneuver. Put a classification onto RNN/LSTM/GRU would be a good choice to get output probabilities. Ryan mentioned using a softmax layer as last layer with nodes equals to the class number.

Also, on time ser

Unlike N-Gram, the current Skip-Gram tokenizer does not allow variable length tokens. This ticket is to implement a min max scheme similar to the N-Gram tokenizer such that each token is a k-skip-n-gram. In other words, instead of fi

Hi @JavierAntoran @stratisMarkou,

First of all, thanks for making all of this code available - it's been great to look through!

Im currently spending some time trying to work through the Weight Uncertainty in Neural Networks in order to implement Bayes-by-Backprop. I was struggling to understand the difference between your implementation of `Bayes-by-Bac