M
Mykel J. Kochenderfer
Researcher at Stanford University
Publications - 449
Citations - 12534
Mykel J. Kochenderfer is an academic researcher from Stanford University. The author has contributed to research in topics: Computer science & Markov decision process. The author has an hindex of 41, co-authored 388 publications receiving 8215 citations. Previous affiliations of Mykel J. Kochenderfer include Massachusetts Institute of Technology & University of Edinburgh.
Papers
More filters
Proceedings ArticleDOI
Monte Carlo Tree Search for Policy Optimization
Xiaobai Ma,Katherine Driggs-Campbell,Katherine Driggs-Campbell,Zongzhang Zhang,Mykel J. Kochenderfer +4 more
TL;DR: This paper presents a method for policy optimization based on Monte-Carlo tree search and gradient-free optimization, which provides a better exploration-exploitation trade-off through the use of the upper confidence bound heuristic.
Journal ArticleDOI
Efficient Aircraft Rerouting During Commercial Space Launches
TL;DR: This data indicates that during a commercial space launch, the Federal Aviation Administration prohibits air traffic within a large column of airspace around the launch trajectory, and the prohibited air traffic is likely to be large and high-speed.
Proceedings Article
Scalable Identification of Partially Observed Systems with Certainty-Equivalent EM
Kunal Menda,Jean de Becdelièvre,Jayesh K. Gupta,Ilan Kroo,Mykel J. Kochenderfer,Zachary Manchester +5 more
TL;DR: It is empirically shown that the certainty-equivalent approximation to expectation-maximization can be a reliable and scalable approach for high-dimensional deterministic systems, which are common in robotics.
Journal ArticleDOI
Portfolio Construction as Linearly Constrained Separable Optimization
TL;DR: A heuristic algorithm based on the alternating direction method of multipliers (ADMM) is proposed that allows for solve times in tens to hundreds of milliseconds with around 1000 securities and 100 risk factors and demonstrates their effectiveness empirically in realistic tax-aware portfolio construction problems.
Proceedings ArticleDOI
Closed-Loop Planning for Disaster Evacuation with Stochastic Arrivals
TL;DR: Close-loop integer programming techniques are shown to obtain up to 90% of the performance of the optimal MDP policy, and can outperform open-loop approaches by as much as 52%.