What are you using it for? There area lot of different approaches to using it. Have you read Sutton and Barto to start with?

Fristlly, i want ot thank all of you for your response

i need a simple mathematical representation of this algorithm to understand it well

about application will be non-linear dynamic system

i do not have sutton and barto Dr Mark Crowly

can you help me more

Sutton and Barto's original book on Reinforcement Learning has a section on Actor-Critic methods http://webdocs.cs.ualberta.ca/~sutton/book/ebook/node66.html

Althought they have a new second edition in progress you can get on Sutton's website as a pdf.

http://webdocs.cs.ualberta.ca/~sutton/book/the-book.html

I think it has more updated information on Actor-critic methods and direct policy search in general. It's a good place to start anyways.

Thanks for your help

i will check it