Learning Distance Metrics for Multi-Label Classification

Henry Gouk, Bernhard Pfahringer, Michael Cree
; Proceedings of The 8th Asian Conference on Machine Learning, PMLR 63:318-333, 2016.

Abstract

Distance metric learning is a well studied problem in the field of machine learning, where it is typically used to improve the accuracy of instance based learning techniques. In this paper we propose a distance metric learning algorithm that is specialised for multi-label classification tasks, rather than the multiclass setting considered by most work in this area. The method trains an embedder that can transform instances into a feature space where Euclidean distance provides an estimate of the Jaccard distance between the corresponding label vectors. In addition to a linear Mahalanobis style metric, we also present a nonlinear extension that provides a substantial boost in performance. We show that this technique significantly improves upon current approaches for instance based multi-label classification, and also enables interesting data visualisations.

Cite this Paper


BibTeX
@InProceedings{pmlr-v63-Gouk8, title = {Learning Distance Metrics for Multi-Label Classification}, author = {Henry Gouk and Bernhard Pfahringer and Michael Cree}, booktitle = {Proceedings of The 8th Asian Conference on Machine Learning}, pages = {318--333}, year = {2016}, editor = {Robert J. Durrant and Kee-Eung Kim}, volume = {63}, series = {Proceedings of Machine Learning Research}, address = {The University of Waikato, Hamilton, New Zealand}, month = {16--18 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v63/Gouk8.pdf}, url = {http://proceedings.mlr.press/v63/Gouk8.html}, abstract = {Distance metric learning is a well studied problem in the field of machine learning, where it is typically used to improve the accuracy of instance based learning techniques. In this paper we propose a distance metric learning algorithm that is specialised for multi-label classification tasks, rather than the multiclass setting considered by most work in this area. The method trains an embedder that can transform instances into a feature space where Euclidean distance provides an estimate of the Jaccard distance between the corresponding label vectors. In addition to a linear Mahalanobis style metric, we also present a nonlinear extension that provides a substantial boost in performance. We show that this technique significantly improves upon current approaches for instance based multi-label classification, and also enables interesting data visualisations.} }
Endnote
%0 Conference Paper %T Learning Distance Metrics for Multi-Label Classification %A Henry Gouk %A Bernhard Pfahringer %A Michael Cree %B Proceedings of The 8th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2016 %E Robert J. Durrant %E Kee-Eung Kim %F pmlr-v63-Gouk8 %I PMLR %J Proceedings of Machine Learning Research %P 318--333 %U http://proceedings.mlr.press %V 63 %W PMLR %X Distance metric learning is a well studied problem in the field of machine learning, where it is typically used to improve the accuracy of instance based learning techniques. In this paper we propose a distance metric learning algorithm that is specialised for multi-label classification tasks, rather than the multiclass setting considered by most work in this area. The method trains an embedder that can transform instances into a feature space where Euclidean distance provides an estimate of the Jaccard distance between the corresponding label vectors. In addition to a linear Mahalanobis style metric, we also present a nonlinear extension that provides a substantial boost in performance. We show that this technique significantly improves upon current approaches for instance based multi-label classification, and also enables interesting data visualisations.
RIS
TY - CPAPER TI - Learning Distance Metrics for Multi-Label Classification AU - Henry Gouk AU - Bernhard Pfahringer AU - Michael Cree BT - Proceedings of The 8th Asian Conference on Machine Learning PY - 2016/11/20 DA - 2016/11/20 ED - Robert J. Durrant ED - Kee-Eung Kim ID - pmlr-v63-Gouk8 PB - PMLR SP - 318 DP - PMLR EP - 333 L1 - http://proceedings.mlr.press/v63/Gouk8.pdf UR - http://proceedings.mlr.press/v63/Gouk8.html AB - Distance metric learning is a well studied problem in the field of machine learning, where it is typically used to improve the accuracy of instance based learning techniques. In this paper we propose a distance metric learning algorithm that is specialised for multi-label classification tasks, rather than the multiclass setting considered by most work in this area. The method trains an embedder that can transform instances into a feature space where Euclidean distance provides an estimate of the Jaccard distance between the corresponding label vectors. In addition to a linear Mahalanobis style metric, we also present a nonlinear extension that provides a substantial boost in performance. We show that this technique significantly improves upon current approaches for instance based multi-label classification, and also enables interesting data visualisations. ER -
APA
Gouk, H., Pfahringer, B. & Cree, M.. (2016). Learning Distance Metrics for Multi-Label Classification. Proceedings of The 8th Asian Conference on Machine Learning, in PMLR 63:318-333

Related Material