simultaneous prediction for the multiple inputs and multipleoutputs scenario

[dajuguan]

Hi @pmorenoz, I am planning to use your algorithm for a multi-inputs and multi-outputs gaussian process online regression problem. It seems that the multi_output.py demo maps each input with respect to each output. However, I want to map the multi-inputs with the multi-outputs at the same time. I have try to modify your code, but I find that the underlying HetMOGP repo also doesn't provide a similar demo for my problem.
So, I am wondering that is it possible to realize the above problem by modifying your awesome continual multi-task Gaussian processes code? If it does, how can I do it? Thanks!