Opportunity at Air Force Research Laboratory (AFRL)
Big Data Analytics, Optimization and Control Theory
Air Force Institute of Technology
||Wright-Patterson AFB, OH 454337542
|Sritharan, Sivaguru S.
||(937)-255-6565 EXT: 3315
This research focuses on the utilization of statistical and machine learning, sparse methods, non-convex and non-smooth optimization techniques to solve nonlinear time variant problems of importance involving big data to the Air Force. Typical problems can come from control and management of turbulence, estimation of ocean-atmospheric flow parameters, complex C4ISR network operations, unmanned systems operations and path planning, and optimal maneuvering of aerial and space vehicles.
Research includes (1) statistical and machine learning methods; (2) compressed sensing techniques; (3) studying variational and quasi-variational inequalities for optimal stopping and impulse control problems of sequential decision making; (4) exploring the connection between differential games and H-infinity control in the context of infinite dimensional problems in fluid flow control; (5) studying mixed-integer programming problems for unmanned systems path planning and obstacle avoidance, and dynamic network optimization problems; and (6) dynamic programming analysis for feedback control synthesis of infinite dimensional time dependent physical processes of interest to the Air Force, such as low/high speed aerodynamics and plasma dynamics problems arising in particular in hypersonic range.
Vapnik, V. N, The Nature of Statistical Learning Theory, Springer-Verlag, 2000
Havarneanu T, Popa C, Sritharan SS: Advances in Differential Equations. 11(8): 893, 2006
Barbu V, Sritharan SS: Nonlinear Analysis, Theory, Methods and Applications 64: 1018, 2006
Statistical and Machine learning, Support Vector Machines, Sparse Methods, Compressed Sensing, Non-convex optimization; Variational inequalities; Non-smooth analysis; Controllability theory; Observability theory; Nonlinear programming; Dynamic programming; Optimal stopping; Mixed-integer programming;
Open to U.S. citizens
Open to Postdoctoral and Senior applicants