Poorya Mianjy

[Poorya Mianjy @pmianjy]

@bneyshabur @MirendilAI @a16z @kleinerperkins Congrats Behnam! Wishing you and the team all the best!

[Poorya Mianjy @pmianjy]

@n4Cr چطور مطوري

[Poorya Mianjy @pmianjy]

@n4Cr نصيييير

[Poorya Mianjy @pmianjy]

@thserra @BucknellU @bucknellmgmt Congrats!

[Poorya Mianjy @pmianjy]

@FeiziSoheil @umdcs Congrats!

[Poorya Mianjy @pmianjy]

At @icmlconf to present "On Dropout and Nuclear Norm Regularization" https://t.co/0IDMizRPJ8
TL;DR in deep linear regression, biconjugate of induced regularizer due to dropout is ~squared nuclear norm of network. Deeper and narrower nets are more biased towards low-rank solutions

[Poorya Mianjy @pmianjy]

@ryandcotterell @ETH Congrats!

[Poorya Mianjy @pmianjy]

@jhuclsp @emnlp2019 and NeurIPS?

[Poorya Mianjy @pmianjy]

@aminkarbasi @NSF Congrats!

[Poorya Mianjy @pmianjy]

@thserra @BucknellU @bucknellmgmt Congrats Thiago! This is really exciting!

[Poorya Mianjy @pmianjy]

@thserra Same here, I don’t know how we missed it. Definitely next time.

[Poorya Mianjy @pmianjy]

Honored for receiving a best poster award in the Princeton Day of Optimization workshop for our work "On the Implicit Bias of Dropout". The full manuscript can be found here: https://t.co/B1uw5nheat #PrincetonDay18

[Poorya Mianjy @pmianjy]

@akshayrangamani Thanks!

[Poorya Mianjy @pmianjy]

@thserra Congratulations!

[Poorya Mianjy @pmianjy]

Excited that our paper "Streaming Kernel PCA with $\tilde{O(\sqrt{n})$ Random Features" is accepted at #NIPS2018!
https://t.co/PvPT9pQ6Vn

[Poorya Mianjy @pmianjy]

@akshayrangamani 🤣

[Poorya Mianjy @pmianjy]

@thserra @icmlconf @teppercmu @CarnegieMellon @UtahSoC Cool!

[Poorya Mianjy @pmianjy]

The camera ready version of our @icmlconf paper "On the Implicit Bias of Dropout" is online: https://t.co/otucSjNfXa
#icml2018