Gradient-bounded dynamic programming for submodular and concave extensible value functions with probabilistic performance guarantees.
Denis LebedevPaul GoulartKostas MargellosPublished in: Autom. (2022)
Keyphrases
- dynamic programming
- greedy algorithm
- objective function
- piecewise linear
- bayesian networks
- continuous functions
- single machine
- parametric family
- database
- coarse to fine
- energy minimization
- reinforcement learning
- object oriented
- stereo matching
- data model
- high order
- uncertain data
- optimal control
- data types
- convex functions
- forward and backward
- multi agent