Handling classification errors by learning to reject classifications

This work investigates how to improve classification metrics by learning when a classifier has higher chances of misclassification. The rejection technique increases the classification reliability and reduces the costs associated with misclassifications on cost-sensitive scenarios. A scenario of bug...

Full description

Saved in:
Bibliographic Details
Main Authors: Junior, Edson Duarte da Silva, Wainer, Jacques, Ferreira, Alexandre
Format: Online
Language:English
Published: Universidade Estadual de Campinas 2018
Subjects:
Online Access:https://econtents.sbu.unicamp.br/eventos/index.php/pibic/article/view/369
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1864358389126004736
author Junior, Edson Duarte da Silva
Wainer, Jacques
Ferreira, Alexandre
author_facet Junior, Edson Duarte da Silva
Wainer, Jacques
Ferreira, Alexandre
author_sort Junior, Edson Duarte da Silva
collection Portal de Eventos Científicos da UNICAMP
container_reference Revista dos Trabalhos de Iniciação Científica da UNICAMP; n. 26 (2018): Congresso de Iniciação Científica Unicamp
description This work investigates how to improve classification metrics by learning when a classifier has higher chances of misclassification. The rejection technique increases the classification reliability and reduces the costs associated with misclassifications on cost-sensitive scenarios. A scenario of bug triaging classification shows the approach effectiveness where classification accuracy is increased from 67% up to 76%.
first_indexed 2025-10-08T13:50:35Z
format Online
id ojs-article-369
institution Universidade Estadual de Campinas
issn 2596-1969
language eng
last_indexed 2025-10-08T13:50:35Z
publishDate 2018
publisher Universidade Estadual de Campinas
record_format ojs
spelling ojs-article-3692025-09-26T14:32:30Z Handling classification errors by learning to reject classifications Junior, Edson Duarte da Silva Wainer, Jacques Ferreira, Alexandre Machine learning Classification rejection Bug triaging. This work investigates how to improve classification metrics by learning when a classifier has higher chances of misclassification. The rejection technique increases the classification reliability and reduces the costs associated with misclassifications on cost-sensitive scenarios. A scenario of bug triaging classification shows the approach effectiveness where classification accuracy is increased from 67% up to 76%. Universidade Estadual de Campinas 2018-12-17 info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion Artigo de convidado application/pdf https://econtents.sbu.unicamp.br/eventos/index.php/pibic/article/view/369 10.20396/revpibic262018369 Revista dos Trabalhos de Iniciação Científica da UNICAMP; n. 26 (2018): Congresso de Iniciação Científica Unicamp 2596-1969 eng https://econtents.sbu.unicamp.br/eventos/index.php/pibic/article/view/369/352 Copyright (c) 2018 Edson Duarte da Silva Junior, Jacques Wainer http://creativecommons.org/licenses/by/4.0
spellingShingle Junior, Edson Duarte da Silva
Wainer, Jacques
Ferreira, Alexandre
Machine learning
Classification rejection
Bug triaging.
Handling classification errors by learning to reject classifications
title Handling classification errors by learning to reject classifications
title_full Handling classification errors by learning to reject classifications
title_fullStr Handling classification errors by learning to reject classifications
title_full_unstemmed Handling classification errors by learning to reject classifications
title_short Handling classification errors by learning to reject classifications
title_sort handling classification errors by learning to reject classifications
topic Machine learning
Classification rejection
Bug triaging.
url https://econtents.sbu.unicamp.br/eventos/index.php/pibic/article/view/369
work_keys_str_mv AT junioredsonduartedasilva handlingclassificationerrorsbylearningtorejectclassifications
AT wainerjacques handlingclassificationerrorsbylearningtorejectclassifications
AT ferreiraalexandre handlingclassificationerrorsbylearningtorejectclassifications