Title :
New reinforcement algorithm in learning automata
Author_Institution :
ETS de Ingenieros de Telecomunicación, Laboratorio de Ordenadores y Cibernética, Madrid, Spain
Abstract :
The letter presents a new stochastic process and its application to learning automata as a reinforcement scheme, and briefly reports some experimental results concerning the comparison between the new algorithm and the most efficient classical reward-inaction scheme.
Keywords :
stochastic automata; stochastic processes; learning automata; reinforcement algorithm; stochastic automata; stochastic process;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19800432
