Dimitri Bertsekas, whose textbooks and algorithms on dynamic programming and convex optimization became foundational references for reinforcement learning and control systems, died on June 3 at age 83. Colleagues are now posting personal reflections that focus less on new technical details and more on the clarity of his writing and the steady guidance he offered students and early-career researchers.
My teacher and friend Dimitri Bertsekas passed away earlier this month. I wrote about his broad contributions to the field of numerical optimization and his deep impact on my writing and research. https://www.argmin.net/p/a-tribute-to-dimitri-bertsekas
Surya Ganguli recalled serving as an undergraduate TA and noted how Bertsekas made rigorous material feel approachable; Ben Recht described him simply as teacher and friend whose books still sit open on many desks.
His formulations in stochastic control and numerical optimization remain embedded in operations research pipelines and modern learning algorithms, though the community has not yet assembled a full retrospective on downstream usage.
No Digg Deeper questions have been answered for this story yet.
Thanks @beenwrekt for a beautiful tribute to Dmitri Bertsekas. I TA'ed for Bertsekas and Tsitsiklis's probabilistic systems course as an undergrad. I was 20 years old and younger than many of the grad students I was teaching so I was quite nervous on the inside. Bertsekas was so kind and he even told me I was one of his favorite TA's (I am sure he said that to many TAs). But that kindness I experienced from him at such a young age, when I was still unsure and forming my scientific self, stuck with me through so many years. Was very sad to hear of his passing.
My teacher and friend Dimitri Bertsekas passed away earlier this month. I wrote about his broad contributions to the field of numerical optimization and his deep impact on my writing and research. https://www.argmin.net/p/a-tribute-to-dimitri-bertsekas