Understanding GANs: the LQG Setting — mentioned by currymj

Sat, 27 Oct 2018 19:04:02 GMT

You are right that they are very closely related. For variational autoencoders (one baseline technique for this sort of thing) if you make your neural network have one layer without a nonlinearity, and train it, it ends up minimizing the same objective as PCA (i.e. finding the eigenfaces). I believe this is also true of GANs where you similarly restrict the generator and discriminator to be very simple. I bet there is a nonlinear non-NN approach that could perform well, but we may not have the investment in hardware, well-optimized algorithms, etc to train big models fast. edit: here's a…

Paper and discussion history. A mention is not an endorsement.

Understanding GANs: the LQG Setting · HackerPapers discussions

Understanding GANs: the LQG Setting — mentioned by currymj