Markov decision processes with exponentially representable discounting

نویسندگان

  • Yair Carmon
  • Adam Shwartz
چکیده

We generalize the geometric discount of finite discounted cost Markov Decision Processes to “exponentially representable” discount functions, prove existence of optimal policies which are stationary from some time N onward, and provide an algorithm for their computation. Outside this class, optimal “N-stationary” policies in general do not exist.

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

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Eventually-stationary policies for Markov decision models with non-constant discounting

We investigate the existance of simple policies in finite discounted cost Markov Decision Processes, when the discount factor is not constant. We introduce a class called “exponentially representable” discount functions. Within this class we prove existence of optimal policies which are eventually stationary—from some time N onward, and provide an algorithm for their computation. Outside this c...

متن کامل

Variance minimization for constrained discounted continuous-time MDPs with exponentially distributed stopping times

This paper deals with minimization of the variances of the total discounted costs for constrained Continuous-Time Markov Decision Processes (CTMDPs). The costs consist of cumulative costs incurred between jumps and instant costs incurred at jump epochs. We interpret discounting as an exponentially distributed stopping time. According to existing theory, for the expected total discounted costs o...

متن کامل

Genetic Programming as Policy Search in Markov Decision Processes

In this paper, we examine genetic programming as a policy search technique for planning problems representable as Markov Decision Processes. The planning task under consideration is derived from a real-time strategy war game. This problem presents unique challenges for standard genetic programming approaches; despite this, we show that genetic programming produces results competitive with stand...

متن کامل

A Genetic Search In Policy Space For Solving Markov Decision Processes

Markov Decision Processes (MDPs) have been studied extensively in the context of decision making under uncertainty. This paper presents a new methodology for solving MDPs, based on genetic algorithms. In particular, the importance of discounting in the new framework is dealt with and applied to a model problem. Comparison with the policy iteration algorithm from dynamic programming reveals the ...

متن کامل

Fuzzy Perceptive Values for MDPs with Discounting

In this paper, we formulate the fuzzy perceptive model for discounted Markov decision processes in which the perception for transition probabilities is described by fuzzy sets. The optimal expected reward, called a fuzzy perceptive value, is characterized and calculated by a new fuzzy relation. As a numerical example, a machine maintenance problem is considered.

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

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

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:
  • Oper. Res. Lett.

دوره 37  شماره 

صفحات  -

تاریخ انتشار 2009