TY - GEN
T1 - A Compact Representation of Multiscale Dissimilarity Data by Prototype Selection
AU - Plasencia-Calaña, Yenisel
AU - Li, Yan
AU - Duin, Robert P.W.
AU - Orozco-Alzate, Mauricio
AU - Loog, Marco
AU - García-Reyes, Edel
PY - 2017
Y1 - 2017
N2 - Multiscale information provides an opportunity to improve the outcomes of data analysis processes. However, if the multiscale information is not properly summarized in a compact representation, this may lead to problems related to high dimensional data. In addition, in some situations, it is convenient to define dissimilarities directly for the multiscale data obtaining in this way a multiscale dissimilarity representation. When these dissimilarities are specifically designed for the problem, it is even possible that they do not fulfill metric requirements. Therefore, standard statistical analysis techniques may not be easily applicable. We propose a new method to combine non-metric multiscale dissimilarities in a compact representation which is used for classification. The method is based on the extended multiscale dissimilarity space and prototype selection, which allows us to handle the potentially non-metric nature of the dissimilarities and exploit the multiscale information at the same time. This is achieved in such a way that the most informative examples per scale are selected. Experimental results show that the approach is promising since it finds a better trade-off in accuracy and efficiency than its counterpart approaches.
AB - Multiscale information provides an opportunity to improve the outcomes of data analysis processes. However, if the multiscale information is not properly summarized in a compact representation, this may lead to problems related to high dimensional data. In addition, in some situations, it is convenient to define dissimilarities directly for the multiscale data obtaining in this way a multiscale dissimilarity representation. When these dissimilarities are specifically designed for the problem, it is even possible that they do not fulfill metric requirements. Therefore, standard statistical analysis techniques may not be easily applicable. We propose a new method to combine non-metric multiscale dissimilarities in a compact representation which is used for classification. The method is based on the extended multiscale dissimilarity space and prototype selection, which allows us to handle the potentially non-metric nature of the dissimilarities and exploit the multiscale information at the same time. This is achieved in such a way that the most informative examples per scale are selected. Experimental results show that the approach is promising since it finds a better trade-off in accuracy and efficiency than its counterpart approaches.
KW - Extended multiscale dissimilarity space
KW - Genetic algorithms
KW - Multiscale data
KW - Prototype selection
UR - http://www.scopus.com/inward/record.url?scp=85013392794&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-52277-7_19
DO - 10.1007/978-3-319-52277-7_19
M3 - Conference contribution
AN - SCOPUS:85013392794
SN - 978-3-319-52276-0
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 150
EP - 157
BT - Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications
A2 - Beltrán-Castañón, Cesar
A2 - Nyström, Ingela
A2 - Famili, Fazel
PB - Springer
CY - Cham
T2 - CIARP 2016
Y2 - 8 November 2016 through 11 November 2016
ER -