noisydonut

3 ms·

1. 1.▲
   WhyNot: A Python sandbox for decision making in dynamics(github.com)

   5 points by noisydonut 7y ago|0 comments
2. 2.▲
   When Recurrent Models Don't Need to Be Recurrent(offconvex.github.io)

   3 points by noisydonut 8y ago|1 comments
3. 3.▲
   Tensor Methods in Machine Learning(offconvex.org)

   8 points by noisydonut 11y ago|0 comments
4. 4.▲
   The reusable holdout: Preserving validity in adaptive data analysis(googleresearch.blogspot.com)

   6 points by noisydonut 11y ago|0 comments
5. 5.▲
   by noisydonut 12y ago

   Nice tl;dr written by Nobel prize winner Al Roth: http://marketdesigner.blogspot.com/2015/03/reflections-on-pr...
6. 6.▲
   by noisydonut 12y ago

   Google recently put a tool called RAPPOR into production that's partially based on differentially private randomized response for collecting usage statistics about Chrome users. Code is here: https://github.com/google&#
7. 7.▲
   How to break a machine learning competition(blog.mrtz.org)

   2 points by noisydonut 12y ago|0 comments