A semismooth Newton method for a kind of HJB equation.
Hongru XuShuilian XiePublished in: Comput. Math. Appl. (2017)
Keyphrases
- reinforcement learning
- hamilton jacobi bellman
- control problems
- newton method
- fischer burmeister
- optimal control
- variational inequalities
- nonlinear complementarity problem
- superlinear convergence
- gauss newton
- convergence analysis
- quasi newton
- stochastic control
- linear equations
- dynamic programming
- regularized least squares
- global convergence
- optimality conditions
- sensitivity analysis
- linear svm
- nonlinear systems
- nonlinear programming
- quadratic programming
- dynamical systems
- neural network
- control strategy
- image processing