Mathematical Finance, Stochastic Analysis, and Machine Learning Seminar by Łukasz Stettner: On Several Time Inconsistent Stochastic Control Problems

Speaker Łukasz Stettner, Department of Probability Theory and Mathematics of Finance, Institute of Mathematics, Polish Academy of Sciences

 

Title: On Several Time Inconsistent Stochastic Control Problems

 

Abstract: In the talk three long run stochastic control problems that are time inconsistent i.e., problems for which Bellman optimality principle does not work, are solved. The problems are: generally discounted average per unit time reward of controlled discrete time Markov processes or impulse control of continuous time Markov processes and log run behavior of a regular utility function of an additive cost. It is shown that these problems can be unexpectedly reduced to time consistent problems in the sense that optimal controls for suitably defined time consistent problems are also optimal for above mentioned time inconsistent ones. The talk is based on the papers [1-3]

 

[1] L. Stettner, Long run stochastic control problems with general discounting,  Applied Mathematics & Optimization (2024) 89:52,

[2] D. Jelito, Ł. Stettner, Impulse control with generalized discounting, SIAM J. Control Optim. 62 no. 2 (2024), 853--876, 

[3] Ł. Stettner, Certainty equivalent control of discrete time Markov processes with the average reward functional, Systems & Control Letters 181 (2023), 105627.

 

 

Mathematical Finance, Stochastic Analysis, and Machine Learning Seminar

Zoom link