Extreme Learning Machines for Multiclass Classification: Refining Predictions with Gaussian Mixture Models

Emil Eirola, Andrey Gritsenko, Anton Akusok, Kaj-Mikael Björk, Yoan Miche, Dušan Sovilj, Rui Nian, Bo He, Amaury Lendasse

Forskningsoutput: Kapitel i bok/rapport/konferenshandlingKonferensbidragVetenskapligPeer review

7 Citeringar (Scopus)

Sammanfattning

This paper presents an extension of the well-known Extreme Learning Machines (ELMs). The main goal is to provide probabilities as outputs for Multiclass Classification problems. Such information is more useful in practice than traditional crisp classification outputs. In summary, Gaussian Mixture Models are used as post-processing of ELMs. In that context, the proposed global methodology is keeping the advantages of ELMs (low computational time and state of the art performances) and the ability of Gaussian Mixture Models to deal with probabilities. The methodology is tested on 3 toy examples and 3 real datasets. As a result, the global performances of ELMs are slightly improved and the probability outputs are seen to be accurate and useful in practice.
OriginalspråkEngelska
Titel på värdpublikationInternational Work-Conference on Artificial Neural Networks : IWANN 2015: Advances in Computational Intelligence
Antal sidor12
UtgivningsortCham
FörlagSpringer
Utgivningsdatum06.06.2015
Sidor153-164
ISBN (tryckt)978-3-319-19221-5
ISBN (elektroniskt)978-3-319-19222-2
DOI
StatusPublicerad - 06.06.2015
MoE-publikationstypA4 Artikel i en konferenspublikation

Publikationsserier

Namn Lecture Notes in Computer Science (LNCS)
Volym9095

Nyckelord

  • 512 Företagsekonomi

Fingeravtryck

Fördjupa i forskningsämnen för ”Extreme Learning Machines for Multiclass Classification: Refining Predictions with Gaussian Mixture Models”. Tillsammans bildar de ett unikt fingeravtryck.

Citera det här