Abstract
Real-time search provides an attractive framework for intelligent autonomous agents, as it allows us to model an agent's ability to improve its performance through experience. However, the behavior of real-time search agents is far from rational during the learning (convergence) process, in that they fail to balance the efforts to achieve a short-term goal (i.e., to safely arrive at a goal state in the present problem solving trial) and a long-term goal (to find better solutions through repeated trials). As a remedy, we introduce two techniques for controlling the amount of exploration, both overall and per trial. The weighted real-time search reduces the overall amount of exploration and accelerates convergence. It sacrifices admissibility but provides a nontrivial bound on the converged solution cost. The real-time search with upper bounds insures solution quality in each trial when the state space is undirected. These techniques result in a convergence process more stable compared with that of the Learning Real-Time A* algorithm.
| Original language | English |
|---|---|
| Pages (from-to) | 1-41 |
| Number of pages | 41 |
| Journal | Artificial Intelligence |
| Volume | 146 |
| Issue number | 1 |
| DOIs | |
| Publication status | Published - 2003 May 1 |
| Externally published | Yes |
Fingerprint
Keywords
- Adaptive learning
- Convergence process
- Rational agent
- Real-time heuristic search
- Resource-boundedness
ASJC Scopus subject areas
- Language and Linguistics
- Linguistics and Language
- Artificial Intelligence
Cite this
Controlling the learning process of real-time heuristic search. / Shimbo, Masashi; Ishida, Toru.
In: Artificial Intelligence, Vol. 146, No. 1, 01.05.2003, p. 1-41.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Controlling the learning process of real-time heuristic search
AU - Shimbo, Masashi
AU - Ishida, Toru
PY - 2003/5/1
Y1 - 2003/5/1
N2 - Real-time search provides an attractive framework for intelligent autonomous agents, as it allows us to model an agent's ability to improve its performance through experience. However, the behavior of real-time search agents is far from rational during the learning (convergence) process, in that they fail to balance the efforts to achieve a short-term goal (i.e., to safely arrive at a goal state in the present problem solving trial) and a long-term goal (to find better solutions through repeated trials). As a remedy, we introduce two techniques for controlling the amount of exploration, both overall and per trial. The weighted real-time search reduces the overall amount of exploration and accelerates convergence. It sacrifices admissibility but provides a nontrivial bound on the converged solution cost. The real-time search with upper bounds insures solution quality in each trial when the state space is undirected. These techniques result in a convergence process more stable compared with that of the Learning Real-Time A* algorithm.
AB - Real-time search provides an attractive framework for intelligent autonomous agents, as it allows us to model an agent's ability to improve its performance through experience. However, the behavior of real-time search agents is far from rational during the learning (convergence) process, in that they fail to balance the efforts to achieve a short-term goal (i.e., to safely arrive at a goal state in the present problem solving trial) and a long-term goal (to find better solutions through repeated trials). As a remedy, we introduce two techniques for controlling the amount of exploration, both overall and per trial. The weighted real-time search reduces the overall amount of exploration and accelerates convergence. It sacrifices admissibility but provides a nontrivial bound on the converged solution cost. The real-time search with upper bounds insures solution quality in each trial when the state space is undirected. These techniques result in a convergence process more stable compared with that of the Learning Real-Time A* algorithm.
KW - Adaptive learning
KW - Convergence process
KW - Rational agent
KW - Real-time heuristic search
KW - Resource-boundedness
UR - http://www.scopus.com/inward/record.url?scp=0037405586&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0037405586&partnerID=8YFLogxK
U2 - 10.1016/S0004-3702(03)00012-2
DO - 10.1016/S0004-3702(03)00012-2
M3 - Article
VL - 146
SP - 1
EP - 41
JO - Artificial Intelligence
JF - Artificial Intelligence
SN - 0004-3702
IS - 1
ER -