[Feature Request] Implement PILCO (Probabilistic Inference for Learning Control)

[PSXBRosa]

Motivation

Implement PILCO (Probabilistic Inference for Learning Control), as requested in #509.

Solution

It uses Gaussian Processes to model dynamics and analytic moment matching to propagate uncertainty, allowing for direct gradient-based policy optimization.

Alternatives

NA

Additional context

Reference: PILCO: A Model-Based and Data-Efficient Approach to Policy Search

Checklist

• I have checked that there is no similar issue in the repo (required)

[PSXBRosa]

Hey, @vmoens

I wanted to let you know I saw the request for PILCO. Since it didn't have a tracking issue, I opened one just now. I'm planning to work on this implementation; please let me know if it's already being handled elsewhere.

[vmoens]

No one's working on this!
Thanks for volunteering!