Papers
6
Total Citations
529
H-Index
4
About
Marcos G. Todorov is a leading figure in the field of stochastic control, with a primary focus on Markov jump linear systems (MJLS). His work addresses the fundamental challenge of controlling systems that experience abrupt, random changes in their dynamics, a problem central to applications in robotics, finance, and networked control. Todorov’s most significant contribution is his seminal book, *Continuous-Time Markov Jump Linear Systems* (2012), which has garnered over 496 citations and serves as a definitive reference in the area. He has also made pioneering advances in robust control, introducing novel operator-theoretic frameworks to solve H₂ and mixed H₂/H∞ problems for both discrete-time and infinite-state MJLS. Notably, his recent research bridges classical control theory with modern machine learning, as seen in his 2018 paper proposing an online temporal-difference (TD) learning algorithm for MJLS—a first step toward reinforcement learning for stochastic hybrid systems. With additional work on count-based control under unknown transition probabilities, Todorov continues to shape the theoretical and practical frontiers of stochastic control, making him an essential read for researchers interested in robust, adaptive, and learning-based control.
Research Focus
Key Achievements
Top Papers
- 1Continuous-Time Markov Jump Linear Systems496 citations · 2012
- 2A new look at the robust control of discrete-time Markov jump linear systems13 citations · 2015
- 3
- 4Online TD(A) for discrete-time Markov jump linear systems8 citations · 2018
- 5
- 6Some Numerical Examples2 citations · 2012