Feasibility Based Large Margin Nearest Neighbor Metric Learning

Hosseini B, Hammer B (2018)
In: ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning. 219-224.

Konferenzbeitrag | Veröffentlicht | Englisch
 
Download
Es wurden keine Dateien hochgeladen. Nur Publikationsnachweis!
Abstract / Bemerkung
Large margin nearest neighbor (LMNN) is a metric learner which optimizes the performance of the popular kNN classifier. However, its resulting metric relies on pre-selected target neighbors. In this paper, we address the feasibility of LMNN's optimization constraints regarding these target points and introduce a mathematical measure to evaluate the size of the feasible region of the optimization problem. We enhance the optimization framework of LMNN by a weighting scheme which prefers data triplets which yield a larger feasible region. This increases the chances to obtain a good metric as the solution of LMNN's problem. We evaluate the performance of the resulting feasibility-based LMNN algorithm using synthetic and real datasets. The empirical results show an improved accuracy for different types of datasets in comparison to regular LMNN.
Erscheinungsjahr
2018
Titel des Konferenzbandes
ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
Seite(n)
219 - 224
Konferenz
26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN)
Konferenzort
Bruges
Konferenzdatum
2018-04-25 – 2018-04-27
ISBN
978-287-587-047-6
Page URI
https://pub.uni-bielefeld.de/record/2919598

Zitieren

Hosseini B, Hammer B. Feasibility Based Large Margin Nearest Neighbor Metric Learning. In: ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning. 2018: 219-224.
Hosseini, B., & Hammer, B. (2018). Feasibility Based Large Margin Nearest Neighbor Metric Learning. ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, 219-224.
Hosseini, Babak, and Hammer, Barbara. 2018. “Feasibility Based Large Margin Nearest Neighbor Metric Learning”. In ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, 219-224.
Hosseini, B., and Hammer, B. (2018). “Feasibility Based Large Margin Nearest Neighbor Metric Learning” in ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning 219-224.
Hosseini, B., & Hammer, B., 2018. Feasibility Based Large Margin Nearest Neighbor Metric Learning. In ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning. pp. 219-224.
B. Hosseini and B. Hammer, “Feasibility Based Large Margin Nearest Neighbor Metric Learning”, ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, 2018, pp.219-224.
Hosseini, B., Hammer, B.: Feasibility Based Large Margin Nearest Neighbor Metric Learning. ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning. p. 219-224. (2018).
Hosseini, Babak, and Hammer, Barbara. “Feasibility Based Large Margin Nearest Neighbor Metric Learning”. ESANN 2018. Proceedings of 26th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning. 2018. 219-224.
Export

Markieren/ Markierung löschen
Markierte Publikationen

Open Data PUB

Quellen

arXiv: 1610.05710

Suchen in

Google Scholar
ISBN Suche