Roadway to Generative Models
It has all the autoencoders from vanilla autoencoders to variational autoencoders Research project : use it for geenrative machine learning, Recommender system etc Next step; variational autoencoder
GAN is next possible application where the performance of GAN and VAE will be clearly visible.
Adversarial autoencoder will be another technique which can be useful for better posterior.
This techniques are very useful for recommender system.
My idea is to model the user product love relationship using VAE and GAN together to extract advantage of both methods.
This techniques are applied on recommender system .
Gourav