Planning by Incremental Dynamic Programming
نویسنده
چکیده
Planning by Incremental Dynamic Programming Richard S. Sutton GTE Laboratories Incorporated Waltham, MA 02254 [email protected] Abstract This paper presents the basic results and ideas of dynamic programming as they relate most directly to the concerns of planning in AI. These form the theoretical basis for the incremental planning methods used in the integrated architecture Dyna. These incremental planning methods are based on continually updating an evaluation function and the situation-action mapping of a reactive system. Actions are generated by the reactive system and thus involve minimal delay, while the incremental planning process guarantees that the actions and evaluation function will eventually be optimal|no matter how extensive a search is required. These methods are well suited to stochastic tasks and to tasks in which a complete and accurate model is not available. For tasks too large to implement the situation-action mapping as a table, supervised-learning methods must be used, and their capabilities remain a signi cant limitation of the approach.
منابع مشابه
Approximate Incremental Dynamic Analysis Using Reduction of Ground Motion Records
Incremental dynamic analysis (IDA) requires the analysis of the non-linear response history of a structure for an ensemble of ground motions, each scaled to multiple levels of intensity and selected to cover the entire range of structural response. Recognizing that IDA of practical structures is computationally demanding, an approximate procedure based on the reduction of the number of ground m...
متن کاملIncremental Policy Generation for Finite-Horizon DEC-POMDPs
Solving multiagent planning problems modeled as DECPOMDPs is an important challenge. These models are often solved by using dynamic programming, but the high resource usage of current approaches results in limited scalability. To improve the efficiency of dynamic programming algorithms, we propose a new backup algorithm that is based on a reachability analysis of the state space. This method, w...
متن کاملIncremental Constraint-Posting Algorithms in Interleaved Planning and Scheduling
In this paper we examine a collection of related incremental constraint-posting algorithms for temporal planning and for planning with continuous processes. The basis for these algorithms is an incremental version of the Bellman-Ford single-source shortest-path algorithm for consistency checking Simple Temporal Networks (STNs). We extend an existing incremental algorithm for STNs and then proce...
متن کاملDynamic Multi Period Production Planning Problem with Semi Markovian Variable Cost (TECHNICAL NOTE)
This paper develops a method for solving the single product multi-period production-planning problem, in which the production and the inventory costs of each period arc concave and backlogging is not permitted. It is also assumed that the unit variable cost of the production evolves according to a continuous time Markov process. We prove that this production-planning problem can be Stated as a ...
متن کاملThe Effect of Feedback based on Inherent and Incremental Ability Theories on Dynamic Balance in Middle-aged Women
The aim of this study was to examine the effect of inherent and incremental ability theories feedback on dynamic balance in middle-aged women. 29 middle-aged women (age: 50-60) randomly assigned into two groups (inherent ability= 15 subjects, and incremental ability= 14 subjects). Both groups after the dynamic balance pretest (Timed Up and Go) received different instructions feedback. Immediate...
متن کامل