Natural actor–critic algorithms

نویسندگان
چکیده

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Natural Algorithms

This Thesis documents our explorations of simple natural physical systems that exhibit useful, interesting computational effects. We view such physical systems as specialized “computers”/natural–computers. These natural–computers (physical systems), as they evolve in phase space, can be seen as performing a computation or executing a natural algorithm to solve a specific problem. We often consi...

متن کامل

Natural algorithms

We provide further evidence that the study of complex self-organizing systems can benefit from an algorithmic perspective. The subject has been traditionally viewed through the lens of physics and control theory. Using tools typically associated with theoretical computer science, we settle an old question in theoretical ecology: bounding the convergence of bird flocks. We bound the time to reac...

متن کامل

Natural actor-critic algorithms

We present four new reinforcement learning algorithms based on actor–critic, natural-gradient and function-approximation ideas, and we provide their convergence proofs. Actor–critic reinforcement learning methods are online approximations to policy iteration in which the value-function parameters are estimated using temporal difference learning and the policy parameters are updated by stochasti...

متن کامل

Learning Algorithms from Natural Proofs

Based on Håstad’s (1986) circuit lower bounds, Linial, Mansour, and Nisan (1993) gave a quasipolytime learning algorithm for AC0 (constant-depth circuits with AND, OR, and NOT gates), in the PAC model over the uniform distribution. It was an open question to get a learning algorithm (of any kind) for the class of AC0[p] circuits (constant-depth, with AND, OR, NOT, and MODp gates for a prime p)....

متن کامل

Natural Algorithms for Optimisation Problems

Many computational techniques borrow ideas from nature in one way or another. Neural networks imitate the structure of our human brain, genetic algorithms simulate evolution and swarms of insects inspired algorithms for stochastic combinatorial optimisation. These techniques are characterised by inherent parallelism, adaptivity, positive feedback and some element of randomness. This report deta...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Automatica

سال: 2009

ISSN: 0005-1098

DOI: 10.1016/j.automatica.2009.07.008