Title: Habituation detection with Allen-Cahn boundary generation

Authors: Takamasa Koshizen, Motohri Kon, Carson Raynolds, Kazuyuki Aihara

Addresses: Honda Research Institute Japan Co. Ltd., 8-1 Honcho, Wako-shi, Saitama 351-0188, Japan. ' Graduate School of Informatics, Kyoto University, Gokasho, Uji, Kyoto 611-0011, Japan. ' Graduate School of Information Science and Technology, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan. ' Graduate School of Information Science, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan

Abstract: We describe a new habituation system that makes use of Partial Differential Equations (PDEs) to deal with time-inconsistent patterns. With respect to our detection system, the Allen-Cahn (AC) equation is used to model a boundary which evolves over time. As a result, the AC could outperform the traditional techniques even real-world dataset. Thus, it leads to a boundary computation for robust detection system.

Keywords: user modelling; habituation; pattern categorisation; time inconsistency; Allen-Cahn equation; boundary growth; partial differential equations; PDEs; robust detection systems.

DOI: 10.1504/IJAIP.2009.026765

International Journal of Advanced Intelligence Paradigms, 2009 Vol.1 No.4, pp.463 - 487

Published online: 25 Jun 2009 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article