5 problems

Quadratic value-function conjecture for the risk-sensitive control problem
arXiv@arxiv
Quadratic value-function conjecture. We conjecture that $V(t,x)$ is quadratic in $x$ :
Conjectureopendynamic-programming risk-sensitive-control stochastic-control
1
The linear feedback relation conjecture for asymmetric risk-sensitive control
arXiv@arxiv
The linear feedback relation conjecture. The adjoint process and the state process are related by
Conjectureopenrisk-sensitive-control stochastic-control
0
Convergence of finite-horizon risk-sensitive stopping values
arXiv@arxiv
Let $\overline{w}_n$ denote the finite-horizon value function for stopping times bounded by $n\in\mathbb{N}$ , and let $u$ denote the corresponding infinite-horizon value function.…
Conjectureopenoptimal-stopping risk-sensitive-control
1
Global convergence conjecture for risk-sensitive value iteration under a near-monotone condition
arXiv@arxiv
Let $\theta=1$ in the near-monotone assumptions, let $c$ have strictly quadratic growth, and suppose that the structural condition … holds for all $x\in\mathbb{R}^d$ , where…
Conjectureopenrisk-sensitive-control stochastic-control value-iteration
0
Feasibility of sampling-based approximate dynamic programming for CVaR MDPs
arXiv@arxiv
A CVaR Markov decision process has a Bellman equation that is contracting, and let a sampling-based approximate dynamic programming approach be applied to it. Sampling-based approx…
Conjectureopenapproximate-dynamic-programming markov-decision-processes risk-sensitive-control
1

5 problems

Quadratic value-function conjecture for the risk-sensitive control problem

The linear feedback relation conjecture for asymmetric risk-sensitive control

Convergence of finite-horizon risk-sensitive stopping values

Global convergence conjecture for risk-sensitive value iteration under a near-monotone condition

Feasibility of sampling-based approximate dynamic programming for CVaR MDPs