Počet záznamů: 1
On-line mixture-based alternative to logistic regression
- 1.
SYSNO ASEP 0464463 Druh ASEP J - Článek v odborném periodiku Zařazení RIV J - Článek v odborném periodiku Poddruh J Článek ve WOS Název On-line mixture-based alternative to logistic regression Tvůrce(i) Nagy, Ivan (UTIA-B) RID, ORCID
Suzdaleva, Evgenia (UTIA-B) RID, ORCIDCelkový počet autorů 2 Zdroj.dok. Neural Network World. - : Ústav informatiky AV ČR, v. v. i. - ISSN 1210-0552
Roč. 26, č. 5 (2016), s. 417-437Poč.str. 20 s. Forma vydání Online - E Jazyk dok. eng - angličtina Země vyd. CZ - Česká republika Klíč. slova on-line modeling ; on-line logistic regression ; recursive mixture estimation ; data dependent pointer Vědní obor RIV BB - Aplikovaná statistika, operační výzkum CEP GA15-03564S GA ČR - Grantová agentura ČR Institucionální podpora UTIA-B - RVO:67985556 UT WOS 000388307600001 EID SCOPUS 85020286021 DOI 10.14311/NNW.2016.26.024 Anotace The paper deals with a problem of modeling discrete variables depending on continuous variables. This problem is known as the logistic regression estimated by numerical methods. The paper approaches the problem via the recursive Bayesian estimation of mixture models with the purpose of exploring a possibility of constructing the continuous data dependent switching model that should be estimated on-line. Here the model of the discrete variable dependent on continuous data is represented as the model of the mixture pointer dependent on data from mixture components via their parameters, which switch according to the activity of the components. On-line estimation of the data dependent pointer model has a great potential for tasks of clustering and classification. The specific subproblems include (i) the model parameter estimation both of the pointer and of the components obtained during the learning phase, and (ii) prediction of the pointer value during the testing phase. Pracoviště Ústav teorie informace a automatizace Kontakt Markéta Votavová, votavova@utia.cas.cz, Tel.: 266 052 201. Rok sběru 2017
Počet záznamů: 1