IngentaConnect A Multiple Resampling Method for Learning from Imbalanced Data Se...

Authors: Estabrooks A.¹; Jo T.²; Japkowicz N.²

Source: Computational Intelligence, Volume 20, Number 1, February 2004 , pp. 18-36(19)

Abstract:

Resampling methods are commonly used for dealing with the class-imbalance problem. Their advantage over other methods is that they are external and thus, easily transportable. Although such approaches can be very simple to implement, tuning them most effectively is not an easy task. In particular, it is unclear whether oversampling is more effective than undersampling and which oversampling or undersampling rate should be used. This paper presents an experimental study of these questions and concludes that combining different expressions of the resampling approach is an effective solution to the tuning problem. The proposed combination scheme is evaluated on imbalanced subsets of the Reuters-21578 text collection and is shown to be quite effective for these problems.

Keywords: inductive learning; decision trees; class imbalance problem; multiple resampling; text classification

Document Type: Research article

DOI: 10.1111/j.0824-7935.2004.t01-1-00228.x

Affiliations: 1: IBM Toronto Lab, Canada 2: University of Ottawa, Canada

The full text electronic article is available for purchase. You will be able to download the full text electronic article after payment.

$35.56 plus tax

A Multiple Resampling Method for Learning from Imbalanced Data Sets