Classifier of news articles via Machine Learning (word embedding and LSTM).
Natural Language Processing models for document classification tasks.
I used a baseline model via a pre-trained 50 dimensional word-embedding representation and a LSTM model
Data: trained the two models on a BBC archive of 2225 news articles (BBC dataset, available on Kaggle).
The LSTM model reached 89% accuracy on the test set.