Tracking Recurring Concepts with Meta-learners

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216224%3A14330%2F09%3A00037443" target="_blank" >RIV/00216224:14330/09:00037443 - isvavai.cz</a>

Nalezeny alternativní kódy

RIV/00216224:14330/09:00067155

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Tracking Recurring Concepts with Meta-learners

Popis výsledku v původním jazyce

This work address data stream mining from dynamic environments where the distribution underlying the observations may change over time. In these contexts, learning algorithms must be equipped with change detection mechanisms. Several methods have been proposed able to detect and react to concept drift. When a drift is signaled, most of the approaches use a forgetting mechanism, by releasing the current model, and start learning a new decision model. It is not rare for the concepts from history to reappear, for example seasonal changes. In this work we present method that memorizes learnt models and uses meta-learning techniques that characterize the domain of applicability of previous models. The meta-learner can detect re-occurrence of contexts and take pro-active actions by activating previous models. The main benefit of this approach is that proposed meta-learner is capable of selecting similar historical concept, if there is one, without the knowledge of true classes of examples.

Název v anglickém jazyce

Tracking Recurring Concepts with Meta-learners

Popis výsledku anglicky

This work address data stream mining from dynamic environments where the distribution underlying the observations may change over time. In these contexts, learning algorithms must be equipped with change detection mechanisms. Several methods have been proposed able to detect and react to concept drift. When a drift is signaled, most of the approaches use a forgetting mechanism, by releasing the current model, and start learning a new decision model. It is not rare for the concepts from history to reappear, for example seasonal changes. In this work we present method that memorizes learnt models and uses meta-learning techniques that characterize the domain of applicability of previous models. The meta-learner can detect re-occurrence of contexts and take pro-active actions by activating previous models. The main benefit of this approach is that proposed meta-learner is capable of selecting similar historical concept, if there is one, without the knowledge of true classes of examples.

Druh

D - Stať ve sborníku

CEP obor

IN - Informatika

OECD FORD obor

—

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2009

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Progress in Artificial Intelligence

ISBN

978-3-642-04685-8

ISSN

0302-9743

e-ISSN

—

Počet stran výsledku

12

Strana od-do

—

Název nakladatele

Springer Berlin / Heidelberg

Místo vydání

Berlin

Místo konání akce

Aveiro

Datum konání akce

12. 10. 2009

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000273296300035