Počet záznamů: 1
A Robustified Metalearning Procedure for Regression Estimators
- 1.
SYSNO ASEP 0510554 Druh ASEP C - Konferenční příspěvek (mezinárodní konf.) Zařazení RIV D - Článek ve sborníku Název A Robustified Metalearning Procedure for Regression Estimators Tvůrce(i) Kalina, Jan (UIVT-O) RID, SAI, ORCID
Neoral, A. (CZ)Zdroj.dok. The 13th International Days of Statistics and Economics Conference Proceedings. - Slaný : Melandrium, 2019 / Löster T. ; Pavelka T. - ISBN 978-80-87990-18-6 Rozsah stran s. 617-626 Poč.str. 10 s. Forma vydání Online - E Akce International Days of Statistics and Economics /13./ Datum konání 05.09.2019 - 07.09.2019 Místo konání Prague Země CZ - Česká republika Typ akce WRD Jazyk dok. eng - angličtina Země vyd. CZ - Česká republika Klíč. slova model choice ; computational statistics ; robustness ; variable selection Vědní obor RIV IN - Informatika Obor OECD Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8) Institucionální podpora UIVT-O - RVO:67985807 UT WOS 000589182000062 DOI https://doi.org/10.18267/pr.2019.los.186.61 Anotace Metalearning represents a useful methodology for selecting and recommending a suitable algorithm or method for a new dataset exploiting a database of training datasets. While metalearning is potentially beneficial for the analysis of economic data, we must be aware of its instability and sensitivity to outlying measurements (outliers) as well as measurement errors. The aim of this paper is to robustify the metalearning process. First, we prepare some useful theoretical tools exploiting the idea of implicit weighting, inspired by the least weighted squares estimator. These include a robust coefficient of determination, a robust version of mean square error, and a simple rule for outlier detection in linear regression. We perform a metalearning study for recommending the best linear regression estimator for a new dataset (not included in the training database). The prediction of the optimal estimator is learned over a set of 20 real datasets with economic motivation, while the least squares are compared with several (highly) robust estimators. We investigate the effect of variable selection on the metalearning results. If the training as well as validation data are considered after a proper robust variable selection, the metalearning performance is improved remarkably, especially if a robust prediction error is used. Pracoviště Ústav informatiky Kontakt Tereza Šírová, sirova@cs.cas.cz, Tel.: 266 053 800 Rok sběru 2020 Elektronická adresa https://msed.vse.cz/msed_2019/sbornik/toc.html
Počet záznamů: 1