Skip to content

In this work, we consider learning sparse models in large scale setting, where the number of samples and the feature dimension can grow as large as millions or billions. Two immediate issues occur under such challenging scenarios: (i) com- putational cost; (ii) memory overhead.

License

Notifications You must be signed in to change notification settings

dhingratul/Large-Scale-Sparse-Models

Error
Looks like something went wrong!

About

In this work, we consider learning sparse models in large scale setting, where the number of samples and the feature dimension can grow as large as millions or billions. Two immediate issues occur under such challenging scenarios: (i) com- putational cost; (ii) memory overhead.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published