Optimal regularity of extended mean field controls and their piecewise constant approximation

Abstract

We consider the control of McKean-Vlasov dynamics whose coefficients have mean field interactions in the state and control. We show that for a class of linear-convex mean field control problems, the unique optimal open-loop control admits the optimal 1/2-Hölder regularity in time. Consequently, we prove that the value function can be approximated by one with piecewise constant controls and discrete-time state processes arising from Euler-Maruyama time stepping, up to an order 1/2 error, and the optimal control can be approximated up to an order 1/4 error. These results are novel even for the case without mean field interaction.

arXiv commentThe weak convergence of the optimal control approximation has been improved to the strong sense

Cited by in corpus (2)

• A fast iterative PDE-based algorithm for feedback controls of nonsmooth mean-field control problems
• Path regularity of coupled McKean-Vlasov FBSDEs