Discovering predictive ensembles for transfer learning and meta-learning

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21240%2F18%3A00328802" target="_blank" >RIV/68407700:21240/18:00328802 - isvavai.cz</a>

Výsledek na webu

<a href="https://link.springer.com/article/10.1007/s10994-017-5682-0" target="_blank" >https://link.springer.com/article/10.1007/s10994-017-5682-0</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/s10994-017-5682-0" target="_blank" >10.1007/s10994-017-5682-0</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Discovering predictive ensembles for transfer learning and meta-learning

Popis výsledku v původním jazyce

Recent meta-learning approaches are oriented towards algorithm selection, optimization or recommendation of existing algorithms. In this article we show how data-tailored algorithms can be constructed from building blocks on small data sub-samples. Building blocks, typically weak learners, are optimized and evolved into data-tailored hierarchical ensembles. Good-performing algorithms discovered by evolutionary algorithm can be reused on data sets of comparable complexity. Furthermore, these algorithms can be scaled up to model large data sets. We demonstrate how one particular template (simple ensemble of fast sigmoidal regression models) outperforms state-of-the-art approaches on the Airline data set. Evolved hierarchical ensembles can therefore be beneficial as algorithmic building blocks in meta-learning, including meta-learning at scale.

Název v anglickém jazyce

Discovering predictive ensembles for transfer learning and meta-learning

Popis výsledku anglicky

Recent meta-learning approaches are oriented towards algorithm selection, optimization or recommendation of existing algorithms. In this article we show how data-tailored algorithms can be constructed from building blocks on small data sub-samples. Building blocks, typically weak learners, are optimized and evolved into data-tailored hierarchical ensembles. Good-performing algorithms discovered by evolutionary algorithm can be reused on data sets of comparable complexity. Furthermore, these algorithms can be scaled up to model large data sets. We demonstrate how one particular template (simple ensemble of fast sigmoidal regression models) outperforms state-of-the-art approaches on the Airline data set. Evolved hierarchical ensembles can therefore be beneficial as algorithmic building blocks in meta-learning, including meta-learning at scale.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2018

Kód důvěrnosti údajů

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

Název periodika

Machine Learning

ISSN

0885-6125

e-ISSN

1573-0565

Svazek periodika

107

Číslo periodika v rámci svazku

1

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

31

Strana od-do

177-207

Kód UT WoS článku

000419684700007

EID výsledku v databázi Scopus

—