News for hackers. Stuff that matters.
Re:Bayesian Generative Adversarial Networks by mendeza on Thursday November 09, 2017 @09:31PM (#15667505)
Student enrolled in his Bayesian Machine Learning class ORIE 6741. The formulation of adding bayesian inference to deep learning can be applied to CNN's as well. I plan on using his work on Stochastic Gradient Hamiltonian Monte Carlo to do Bayesian Inference on Image Segmentation as well. For more interest, check out this blog, super interesting: http://www.cs.ox.ac.uk/people/yarin.gal/website/blog_3d801aa...
Re:Bayesian Generative Adversarial Networks by tMcGrath on Friday November 10, 2017 @04:55AM (#15668928)
I'd like to flag up this paper on HMC and minibatch inference: https://arxiv.org/abs/1502.01510v1. It seems that SGHMC might not always behave as well as you might like.

I should say this isn't to pour cold water on the idea of Bayesian deep learning - I think it's an extremely sensible idea and undoubtedly a good direction for research, but just wanted to caution that this particular MCMC method might have subtle flaws.

Re:Bayesian Generative Adversarial Networks by eref on Friday November 10, 2017 @03:09PM (#15672546)
I wished more academic papers would have a "Limitations" section. Often it is so hard to tell what these are, e.g.: Does it scale well? How much faster/slower is it than other methods? Does it converge reliably?
Re:Bayesian Generative Adversarial Networks by pishpash on Friday November 10, 2017 @06:48PM (#15674074)
No incentive for such a thing.
Re:Bayesian Generative Adversarial Networks by eref on Friday November 10, 2017 @07:58PM (#15674379)
Neither is there for reproducible code, but now there are artificial incentives such as the Reproduction Challenge.
Re:Bayesian Generative Adversarial Networks by MrQuincle on Friday November 10, 2017 @02:58AM (#15668600)
Did you read "Variational Gaussian Dropout is not Bayesian", https://arxiv.org/abs/1711.02989?

https://twitter.com/dustinvtran/status/928493076854669312

Re:Bayesian Generative Adversarial Networks by elcritch on Friday November 10, 2017 @02:54PM (#15672430)
I posted before refreshing and missed the link to this paper. Will need to read through it.
Re:Bayesian Generative Adversarial Networks by elcritch on Friday November 10, 2017 @02:51PM (#15672404)
Interesting blog post! From a 10,000 ft overview is this method utilizing the conditional posterior's to incorporate uncertainty information into the model GAN model? If so how does it differ from using drop-out's to approximate Bayesian behavior ala https://arxiv.org/pdf/1506.02142.pdf ?