Integer Self-Organizing Maps for Digital Hardware

ISBN för värdpublikation: 978-1-7281-1985-4
The Self-Organizing Map algorithm has been proven and demonstrated to be a useful paradigm for unsupervised machine learning of two-dimensional projections of multidimensional data. The tri-state Self-Organizing Maps have been proposed as an accelerated resource-efficient alternative to the Self-Organizing Maps for implementation on field-programmable gate array (FPGA) hardware. This paper presents a generalization of the tri-state Self-Organizing Maps. The proposed generalization, which we call integer Self-Organizing Maps, requires only integer operations for weight updates. The presented experiments demonstrated that the integer Self-Organizing Maps achieve better accuracy in a classification task when compared to the original tri-state Self-Organizing Maps.

Osipov, EvgenyLuleå tekniska universitet,Datavetenskap(Swepub:ltu)eao (författare)
De Silva, DaswinLa Trobe University, Melbourne, Australia (författare)
Wiklund, UrbanUmeå universitet,Radiofysik,Umeå University, Umeå, Sweden(Swepub:umu)urwi0001 (författare)
Alahakoon, DammindaLa Trobe University, Melbourne, Australia (författare)
Luleå tekniska universitetDatavetenskap (creator_code:org_t)

Ingår i:2019 International Joint Conference on Neural Networks (IJCNN): IEEE9781728119854

Av författaren/redakt...: Kleyko, Denis, 1 ...; Osipov, Evgeny; De Silva, Daswin; Wiklund, Urban; Alahakoon, Dammi ...

Om ämnet

Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.