Can we use autoencoders or variational autoencoders for function approximation?

As far as I checked the net the autoencoders are used mostly in image or audio application but the question is that can we use them for fitting models?

05 June 2018 5,723 3 View