Predictions for Online News Popularity

Barcelona GSE Master in Data Science

Kaggle Competition https://inclass.kaggle.com/c/predicting-online-news-popularity

Roger Cusco, Matthew Sudmann-Day, Miquel Torrens

This github repository contains the source code, data, documentation, results, and an R package for the predictions on the Kaggle competition performed by this team.

The structure of the repository contains the following folders:

• data: RData files containing intermediate results of the analyses
• input: original input files downloaded from the website
• output: CSV files with some results of the models handed in
• package: files related to the R package built for the project
• syntax: source code with the R scripts used to produce the results
• temp: temporary files used and stored during the development of the project

The final report can be found in the doc folder.