Research areas
- Risk-sensitive stochastic control — particularly extensions of Jacobson-style LEQG analysis into settings where the standard exponential transformation fails (Markov jump systems, partial information, mean-field).
- Markov jump linear systems (MJLS) — optimal control, filtering, and robustness for regime-switching linear systems.
- H∞ and minimax control — robustness analysis in the Başar-Bernhard tradition.
- Mean-field games and control — collaboration thread with Başar.
Key papers
- moon-basar-risk-sensitive-control-mjls (2016) — co-author with Tamer Başar on the MJLS risk-sensitive control impossibility result.
Recent work
(Pending further ingest.)
Collaborators
- Tamer Başar (tamer-basar) — PhD-era and follow-up co-author at UIUC.
My notes
In the CRE asset-pricing project, Moon-Başar 2016 is the relevant reference for any future attempt to robustify the MSRE / RBPF pipeline via risk-sensitive substitution — and the answer is “you cannot do this for free in MJLS.” Importance ~4: not yet a household name in the broader ML/control community, but a load-bearing co-author on a paper that closes a structural door for an entire family of would-be robustification schemes.