The Forest Through the Trees in Multilabel Classification

Tuesday, May 7, 2013 at 1:00pm to 1:45pm

Bronfman Science Center, 106 18 Hoxsey St, Williamstown, MA 01267, USA

The Forest Through the Trees in Multilabel Classification

Benjamin Seiler '13

Mathematics and Statistics Department Thesis Defense

Abstract:  Traditional machine learning classification algorithms are not suited for statistical classification problems in which an instance can simultaneously belong to more than one class.  Such multilabel classification problems have prompted significant research in recent years including a concerted effort to bridge the gap between established classification techniques and this nonstandard framework.  Based on such works as recently as Tsoumakas and Katakis [2007] and Vogrincic and Bosnic [2011], the vast majority of novel multilabel classification algorithms are compared to baseline problem transformation techniques using only support vector machines or linear models.  In this study, we broaden the pool of potential base learners for problem transformation techniques and discover significant evidence to suggest the superiority of partition tree based methods in many cases, thereby, raising the bar for baseline competitiveness.

Event Type



Mathematics & Statistics

Recent Activity

People Going

Getting Here