You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Predicting Popular Government, an analys is of states
By Justin Hines
This is a final project for Data Driven Modeling taught by Jake Hofman.
This project aims to predict the popularity of governemntal articles using a multi-class bernoulli naive bayes classifier.
Run make to download the dataset, filter, and apply the classifier. Results will be stored in log/. All tsv data sets will be in tsv organized by dstate. Final filtered tsvs will be in tsv/{STATE}/tsvs/
Run make bayes to run the bayes classifier on a dataset, or python bayes.py in the src/ directory.