Please use this identifier to cite or link to this item: http://repositorio.unicamp.br/jspui/handle/REPOSIP/330115
Type: Artigo
Title: Nearest Neighbors Distance Ratio Open-set Classifier
Nearest neighbors distance ratio open-set classifier
Author: Mendes Junior, Pedro R.
Souza, Roberto M. de
Weneck, Rafael de O.
Stein, Bernardo V.
Pazinato, Daniel V.
Almeida, Waldir R. de
Penatti, Otávio A. B.
Torres, Ricardo da S.
Rocha, Anderson
Abstract: In this paper, we propose a novel multiclass classifier for the open-set recognition scenario. This scenario is the one in which there are no a priori training samples for some classes that might appear during testing. Usually, many applications are inherently open set. Consequently, successful closed-set solutions in the literature are not always suitable for real-world recognition problems. The proposed open-set classifier extends upon the Nearest-Neighbor (NN) classifier. Nearest neighbors are simple, parameter independent, multiclass, and widely used for closed-set problems. The proposed Open-Set NN (OSNN) method incorporates the ability of recognizing samples belonging to classes that are unknown at training time, being suitable for open-set recognition. In addition, we explore evaluation measures for open-set problems, properly measuring the resilience of methods to unknown classes during testing. For validation, we consider large freely-available benchmarks with different open-set recognition regimes and demonstrate that the proposed OSNN significantly outperforms their counterparts in the literature.
In this paper, we propose a novel multiclass classifier for the open-set recognition scenario. This scenario is the one in which there are no a priori training samples for some classes that might appear during testing. Usually, many applications are inher
Subject: Open-set Recognition
Nearest Neighbor Classifier
Open-set Nearest-neighbor Classifier
Nearest Neighbors Distance Ratio
Open-set Evaluation Measures
Reconhecimento em cenário aberto
Classificação multi-classe
Reconhecimento de padrões
Inteligência artificial
Country: Estados Unidos
Editor: Springer
Citation: Machine Learning. Springer, v. 106, p. 359 - 386, 2017.
Rights: fechado
Aberto
Identifier DOI: 10.1007/s10994-016-5610-8
Address: https://link.springer.com/article/10.1007/s10994-016-5610-8
Date Issue: 2017
Appears in Collections:IC - Artigos e Outros Documentos
FEA - Artigos e Outros Documentos
FEEC - Artigos e Outros Documentos

Files in This Item:
File SizeFormat 
000394355700002.pdf4.85 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.