COMP 5005
Learning Systems for Random Environments
Computerized adaptive learning for random environments and its applications. Topics include a mathematical review, learning automata which are deterministic/stochastic, with fixed/variable structures, of continuous/discretized design, with ergodic/absorbing properties and of estimator families.
2023
Topics
- random environments.
- and difference equations.
- Material: The heart of the course will involve deterministic and stochastic learning
- algorithms studied will be the linear and nonlinear learning schemes of the
- continuous and discretized families with ergodic and non-ergodic properties.
- Applications: Applications of learning automata in file allocation, game playing, path
- finding, optimization, solving knapsack problems and in decision making
- will be discussed.
- Mails
- Copying
Readings
- K. S. Narendra and M. A. L. Thathachar Learning Automata, Prentice-Hall, 1989 (or later).