On the policy improvement algorithm for ergodic risk-sensitive control

Abstract

In this article we consider the ergodic risk-sensitive control problem for a large class of multidimensional controlled diffusions on the whole space. We study the minimization and maximization problems under either a blanket stability hypothesis, or a near-monotone assumption on the running cost. We establish the convergence of the policy improvement algorithm for these models. We also present a more general result concerning the region of attraction of the equilibrium of the algorithm.

arXiv comment20 pages

References in corpus (8)

• Principal eigenvalues and an anti-maximum principle for homogeneous fully nonlinear elliptic equations
• On the structure of solutions of ergodic type Bellman equation related to risk-sensitive control
• Strict monotonicity of principal eigenvalues of elliptic operators in ${\mathbb{R}}^d$ and risk-sensitive control
• Infinite horizon risk-sensitive control of diffusions without any blanket stability assumptions
• A variational formula for risk-sensitive control of diffusions in ${\mathbb{R}}^d$
• Ergodic control of a class of jump diffusions with finite Lévy measures and rough kernels
• Yet again on iteration improvement for averaged expected cost control for 1D ergodic diffusions
• Risk-sensitive control for a class of diffusions with jumps

Cited by in corpus (1)

• Ergodic Risk-Sensitive Control of Markov Processes on Countable State Space Revisited