Samarth Bhargav

Explore the methodology and intricacies of Twitter's algorithm

1

We should try to replicate the algorithm in base R code to understand how it works, and how to select parameters. The output should be a R notebook.

Right now, everything is in `src` We have to break this down into folders based on function, like `redux`, `components`, etc.

The current data doesn't show any anomalies. Add random noise with a configurable probability.

Ref: - [1] - http://www.jmlr.org/proceedings/papers/v39/oliveira14.pdf - [2] - http://stats.stackexchange.com/questions/47950/ensemble-time-series-model