Start Globally, Optimize Locally, Predict Globally: Improving Performance on Imbalanced Data
Authors:
- David A. CieslakDCDavid A. Cieslak
- Nitesh V. ChawlaNCNitesh V. Chawla
- 1 December 2008
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15504786,p. 143-152
Abstract
Class imbalance is a ubiquitous problem in supervised learning and has gained wide-scale attention in the literature. Perhaps the most prevalent solution is to apply sampling to training data in order improve classifier performance. The typical approach will apply uniform levels of sampling globally. However, we believe that data is typically multi-modal, which suggests sampling should be treated locally rather than globally. It is the purpose of this paper to propose a framework which first identifies meaningful regions of data and then proceeds to find optimal sampling levels within each. This paper demonstrates that a global classifier trained on data locally sampled produces superior rank-orderings on a wide range of real-world and artificial datasets as compared to contemporary global sampling methods.Keywords
This publication has 15 references indexed in Scilit:
- A framework for monitoring classifiers’ performance: when and why failure occurs?Knowledge and Information Systems, 2008
- Automatically countering imbalance and its empirical relationship to costData Mining and Knowledge Discovery, 2008
- Local decomposition for rare class analysis2007
- Classification and knowledge discovery in protein databasesJournal of Biomedical Informatics, 2004
- A study of the behavior of several methods for balancing machine learning training dataACM SIGKDD Explorations Newsletter, 2004
- Mining with rarityACM SIGKDD Explorations Newsletter, 2004
- EditorialACM SIGKDD Explorations Newsletter, 2004
- Combined 5 × 2 cv F Test for Comparing Supervised Classification Learning AlgorithmsNeural Computation, 1999
- Approximate Statistical Tests for Comparing Supervised Classification Learning AlgorithmsNeural Computation, 1998
- Induction of decision treesMachine Learning, 1986