Self-Organizing Map for Temporal Signal Processing and Its Application: An Analysis of Human Decision-Making Style

Accession number;05A0243387
Title;Self-Organizing Map for Temporal Signal Processing and Its Application: An Analysis of Human Decision-Making Style
Author; WAKUYA HIROSHI (Saga Univ., Fac. Sci. and Engineering, JPN) TANIGAWA MASAHIRO (Saga Univ., Fac. Sci. and Engineering, JPN)
Journal Title;Reports of the Faculty of Science and Engineering, Saga University
Journal Code:G0861A
ISSN:0385-6186
VOL.33;NO.2;PAGE.1-9(2004)
Figure&Table&Reference;FIG.6, TBL.5, REF.10
Pub. Country;Japan
Language;Japanese
Abstract;Recently, a new architecture of the self-organizing map (SOM) for temporal signal processing is undertaken by a lot of researchers. In order to evaluate its performance, a Janken game is adopted in this study. Generally speaking, a game is repetition of decision-making process, i.e., the current action is often determined based on one's experience, so it can be seen as a sort of temporal signal processing. Then, 13 people are chosen and matched each other, and each recorded sequence of their hand-signs (rock, scissors and paper) is analyzed. As a result, it is found experimentally that the trained SOM can develop feature maps which reflect subject's personalities successfully. (author abst.)