A sparse code increases the speed and efficiency of neuro-dynamic programming for optimal control tasks with correlated inputs.
Peter N. LoxleyPublished in: Neurocomputing (2021)
Keyphrases
- optimal control
- dynamic programming
- control problems
- optimal control problems
- risk sensitive
- feedback control
- class of nonlinear systems
- infinite horizon
- control strategy
- brownian motion
- reinforcement learning
- greedy algorithm
- multistage
- lyapunov function
- neural network
- control law
- knapsack problem
- linear programming
- artificial neural networks
- dea model
- neuro fuzzy
- control theory
- real time