Cheng Ju

PhD@Berkeley

[Cheng Ju @StatCheng]

Excited to announce that my work with David Benkeser and Mark van der Laan @laan12345 on outcome-highly-adaptive-lasso has been published by Biometrics! https://t.co/6F9KvSSluE

[Manu Kumar @manukumardatyal]

Path for Ph.D.

[Cheng Ju @StatCheng]

Nice work https://t.co/87T8DDFB0g

[Little Boat @Little__Boat]

NN models used to win recent Avito kaggle competition.

[Lihua Lei @lihua_lei_stat]

Here is our spotlight video https://t.co/ZI38fwNyRB on non-convex SCSG #nips2017, the first algorithm that is provably strictly better than SGD in all regimes, with Cheng Ju, Jianbo Chen and Michael Jordan. Paper: https://t.co/Wbh7po7Ge0

[Cheng Ju @StatCheng]

We are delighted to announce the first official R package ‘ctmle’ https://t.co/7AGoWGe1Vw for collaborative targeted learning. It implements the vanilla/scalable variable selection C-TMLE, the glmnet-CTMLE, and a general template for exposure model selection @target_learning.