Generalised Discount Functions
نویسندگان
چکیده
In recent years, work has been done to develop the theory of General Reinforcement Learning (GRL). However, there are no examples demonstrating the known results regarding generalised discounting. We have added to the GRL simulation platform (AIXIjs) the functionality to assign an agent arbitrary discount functions, and an environment which can be used to determine the effect of discounting on an agent’s policy. Using this, we investigate how geometric, hyperbolic and power discounting affect an informed agent in a simple MDP. We experimentally reproduce a number of theoretical results, and discuss some related subtleties. It was found that the agent’s behaviour followed what is expected theoretically, assuming appropriate parameters were chosen for the Monte-Carlo Tree Search (MCTS) planning algorithm.
منابع مشابه
Generalised Discount Functions applied to a Monte-Carlo AI u Implementation
In recent years, work has been done to develop the theory of General Reinforcement Learning (GRL). However, there are few examples demonstrating the known results regarding generalised discounting. We have added to the GRL simulation platform AIXIjs the functionality to assign an agent arbitrary discount functions, and an environment which can be used to determine the effect of discounting on a...
متن کاملGeneralized Ritt type and generalized Ritt weak type connected growth properties of entire functions represented by vector valued Dirichlet series
In this paper, we introduce the idea of generalized Ritt type and generalised Ritt weak type of entire functions represented by a vector valued Dirichlet series. Hence, we study some growth properties of two entire functions represented by a vector valued Dirichlet series on the basis of generalized Ritt type and generalised Ritt weak type.
متن کاملCritical properties of the double - frequency sine - Gordon model with applications
We study the properties of the double-frequency sine–Gordon model in the vicinity of the Ising quantum phase transition displayed by this model. Using a mapping onto a generalised lattice quantum Ashkin-Teller model, we obtain critical and nearly-off-critical correlation functions of various operators. We discuss applications of the double-sine-Gordon model to one-dimensional physical systems, ...
متن کاملMeasuring Impatience in Intertemporal Choice
In general terms, decreasing impatience means decreasing discount rates. This property has been usually referred to as hyperbolic discounting, although there are other discount functions which also exhibit decreasing discount rates. This paper focuses on the measurement of the impatience associated with a discount function with the aim of establishing a methodology to compare this characteristi...
متن کاملMarkov Decision Processes with General Discount Functions
In Markov Decision Processes, the discount function determines how much the reward for each point in time adds to the value of the process, and thus deeply a ects the optimal policy. Two cases of discount functions are well known and analyzed. The rst is no discounting at all, which correspond to the totaland average-reward criteria. The second case is a constant discount rate, which leads to a...
متن کامل