Title :
Neural networks for classification and recognition of individual users in adaptive human-computer interface
Author :
Dimitrova, M. ; Hubert, M. ; Boyadjiev, D. ; Nabout, A. ; Zaprianov, N.J. ; Eldin, H. A Nour
Author_Institution :
Inst. of Control & Syst. Res., Bulgarian Acad. of Sci., Sofia, Bulgaria
Abstract :
The paper reports on a methodology for the identification of novel users from responses to computer-generated tasks. It is proposed that the cluster structure of the initial pool of data is identified. Standard neural algorithms are employed for classification according to the cluster structure. For recognition of patterns of data within each class, a compound network is employed that “individualizes” the patterns and stores them for subsequent recognition
Keywords :
adaptive systems; biometrics (access control); neural nets; pattern classification; user modelling; adaptive human-computer interface; compound network; computer-generated tasks; data cluster structure identification; data pattern recognition; neural algorithms; neural networks; pattern individualization; user classification; user recognition; user responses; Artificial intelligence; Automatic generation control; Computer interfaces; Computerized monitoring; Control systems; Humans; Intelligent networks; Neural networks; Pattern recognition; Psychology;
Conference_Titel :
Intelligent Control, 1997. Proceedings of the 1997 IEEE International Symposium on
Conference_Location :
Istanbul
Print_ISBN :
0-7803-4116-3
DOI :
10.1109/ISIC.1997.626421
