Modelling - opt-out-tools/opt-out GitHub Wiki

Features

• Retweet count
• URL's
• Emoji's
• Conversation history
• Sentiment
• word2vec
• TF-IDF

Binary Classifiers to evaluate:

• SVM
• GBM
• Deep Neural Net
• Recurrent Neural Net
• Logistic Regression
• Naive Bayes

Modelling Considerations

• cross-platform data set issues. Will a data set from one platform perform on another. Should calls to the API be platform specific.

Modelling Resources

• https://www.kaggle.com/c/jigsaw-toxic-comment-classification-challenge/kernels