Published in: 12th Iranian Conference on Electric Engineering
COI code: ICEE12_223
Paper Language: English
How to Download This Paper
For Downloading the Fulltext of CIVILICA papers please visit the orginal Persian Section of website.
Authors Multi-Step Action Selection in Hunter Prey ProblemM. B. Naghibi -S - Ferdowsi University of Mashhad
M. R. Akbarzadeh -T - Ferdowsi University of Mashhad
Abstract:Q-Learning is a popular agent learning algorithm but it has several weaknesses such as slow convergence in large state space.
Generalization methods that try to reduce the size of state space may produce a perceptual aliasing problem. In this paper we use only two states for a hunter that tries to catch a random or intelligent prey in a 10*10 square domain. We show that, because of perceptual aliasing, random action selection fairs better than the strategies found by Q-learning when prey acts randomly. Furthermore
a novel multi-step action selection technique is introduced to decrease the exploration steps that hunter needs to catch the prey. Results show that the proposed algorithm improves number of actions taken for catching both intelligent and random preys by over 50%.
Keywords:Reinforcement Learning, Q-learning, Agent Behavior, Perceptual Aliasing, Hunter Prey Problem
COI code: ICEE12_223
how to cite to this paper:If you want to refer to this article in your research, you can easily use the following in the resources and references section:
Naghibi -S, M. B. & M. R. Akbarzadeh -T, 2004, Multi-Step Action Selection in Hunter Prey Problem, 12th Iranian Conference on Electric Engineering, مشهد, دانشگاه فردوسي مشهد, https://www.civilica.com/Paper-ICEE12-ICEE12_223.htmlInside the text, wherever referred to or an achievement of this article is mentioned, after mentioning the article, inside the parental, the following specifications are written.
First Time: (Naghibi -S, M. B. & M. R. Akbarzadeh -T, 2004)
Second and more: (Naghibi -S & Akbarzadeh -T, 2004)
For a complete overview of how to citation please review the following CIVILICA Guide (Citation)
The University/Research Center Information:
Type: state university
Paper No.: 25310
in University Ranking and Scientometrics the Iranian universities and research centers are evaluated based on scientific papers.
Research Info Management
Export Citation info of this paper to research management softwares
New Related Papers
- A Hybrid Islanding Detection Technique Using Rate of Voltage and Frequency Change and Reactive Power Compensation
- Online Voltage Instability Prediction Using Novel Indices from Trajectory of Voltage Profile
- Blood Glucose Control Algorithms for Diabetics
- Single PhasePhotovoltaic System Connection to Grid Based OntheZ-SourceInverter
- Reconfiguration and Capacitor Allocation in Distribution Systems to Reduce Power Losses and Improve Voltage Profiles using Ant Lion Algorithm
The Above articles are recently indexed in the related subjects
Iran Scientific Advertisment Netword
Share this paper
WHAT IS COI?
COI is a national code dedicated to all Iranian Conference and Journal Papers. the COI of each paper can be verified online.