Skip to content

Online Bayesian Learning with Non-Uniform Shrinkage to Improve Feature Selection for Click-Through-Rate Prediction and Text Classification

Notifications You must be signed in to change notification settings

zsdlightning/OLSS

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 

Repository files navigation

MIT License

This poject is the source code for the paper, "Online Bayesian Feature Selection with Non-Uniform Shrinkage for CTR Prediction and Text Classification" by Shandian Zhe (http://www.cs.utah.edu/~zhe/), et. al.

Due to the data privacy issues, we cannot release the data/code related to Yahoo! Ads platforms. However, we publish the code for the text classfication tasks, including Rcv1 and news20.

The data is too big to be uploaded into github. The data can be downloaded from http://www.cs.utah.edu/~zhe/data/olss-data.tgz

Note the "vw" folder contains the executable vw software in Linux. This can also be downloaded from http://finance.yendor.com/ML/VW/Binaries/

To run the code, clone the project. Then download and extract the data in parallel to the folder "OLSS". Follow README to run the code in OLSS folder.

About

Online Bayesian Learning with Non-Uniform Shrinkage to Improve Feature Selection for Click-Through-Rate Prediction and Text Classification

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published