نتایج جستجو برای: bellman zadehs principle
تعداد نتایج: 157398 فیلتر نتایج به سال:
This paper investigates possible methods of parallelizing two common single source path-finding algorithms. This research could enable faster computation of optimal path planning in complex environments. Both algorithms, A* and Bellman-Ford were found to be parallelizable. After testing, A* was shown to lack good scaling potential while Bellman-Ford showed promise as a scalable parallel path fi...
Several recent papers have proposed recursive Lagrangian-basedmethods for solving dynamic contracting problems. Thesemethods give rise to Bellman operators that incorporate either a dual inf-sup or a saddle point operation. We give conditions that ensure the Bellman operator implied by a dual recursive formulation is contractive. JEL codes: C61, C73, D82, E61.
We consider approximation methods for Markov decision processes in the learning and simulation context. For policy evaluation based on solving approximate versions of a Bellman equation, we propose the use of weighted Bellman mappings. Such mappings comprise weighted sums of one-step and multistep Bellman mappings, where the weights depend on both the step and the state. For projected versions ...
This paper studies the infinite-horizon optimal consumption problem with a path-dependent reference under exponential utility. The performance is measured by difference between nonnegative rate and fraction of historical maximum. running maximum process chosen as an auxiliary state process, hence value function depends on two variables. Hamilton–Jacobi–Bellman (HJB) equation can be heuristicall...
We study existence and uniqueness of a fixed point for the Bellman operator in deterministic dynamic programming. Without any topological assumption, we show that the Bellman operator has a unique fixed point in a restricted domain, that this fixed point is the value function, and that the value function can be computed by value iteration.
Discrete inequalities, in particular the discrete analogues of the Gronwall–Bellman inequality, have been extensively used in the analysis of finite difference equations. The aim of the present paper is to establish some fractional difference inequalities of Gronwall–Bellman type which provide explicit bounds for the solutions of fractional difference equations.
We obtain error bounds for monotone approximation schemes of Hamilton-Jacobi-Bellman equations. These bounds improve previous results of Krylov and the authors. The key step in the proof of these new estimates is the introduction of a switching system which allows the construction of approximate, (almost) smooth supersolutions for the Hamilton-Jacobi-Bellman equation.
We derive error estimates for certain approximate solutions of Bellman equations associated to a class of controlled jump-diffusion (Lévy) processes. These Bellman equations are fully nonlinear degenerate integroPDEs interpreted in the sense of viscosity solutions. The approximate solutions are generated by an implicit finite difference-quadrature scheme.
Temporal-diierence (TD) learning can be used not just to predict rewards, as is commonly done in reinforcement learning, but also to predict states, i.e., to learn a model of the world's dynamics. We present theory and algorithms for intermixing TD models of the world at diierent levels of temporal abstraction within a single structure. Such multi-scale TD models can be used in model-based rein...
There is growing interest in the use of grid–level storage to smooth variations in supply that are likely to arise with increased use of wind and solar energy. Energy arbitrage, the process of buying, storing, and selling electricity to exploit variations in electricity spot prices, is becoming an important way of paying for expensive investments into grid level storage. Independent system oper...
نمودار تعداد نتایج جستجو در هر سال
با کلیک روی نمودار نتایج را به سال انتشار فیلتر کنید