The benefits of using multi-objectivization for mining Pittsburgh partial classification rules in imbalanced and discrete data - Archive ouverte HAL Access content directly
Conference Papers Year : 2013

The benefits of using multi-objectivization for mining Pittsburgh partial classification rules in imbalanced and discrete data

Abstract

A large number of rule interestingness measures have been used as objectives in multi-objective classification rule mining algorithms. Aggregation or Pareto dominance are commonly used to deal with these multiple objectives. This paper compares these approaches on a partial classification problem over discrete and imbalanced data. After performing a Principal Component Analysis (PCA) to select candidate objectives and find conflictive ones, the two approaches are evaluated. The Pareto dominance-based approach is implemented as a dominance-based local search (DMLS) algorithm using confidence and sensitivity as objectives, while the other is implemented as a single-objective hill climbing using F-Measure as an objective, which combines confidence and sensitivity. Results shows that the dominance-based approach obtains statistically better results than the single-objective approach.
Not file

Dates and versions

hal-00806777 , version 1 (02-04-2013)

Identifiers

  • HAL Id : hal-00806777 , version 1

Cite

Julie Jacques, Julien Taillard, David Delerue, Laetitia Jourdan, Clarisse Dhaenens. The benefits of using multi-objectivization for mining Pittsburgh partial classification rules in imbalanced and discrete data. Genetic and Evolutionary Computation Conference (GECCO 2013), Jul 2013, Amsterdam, Netherlands. pp.543-550. ⟨hal-00806777⟩
245 View
0 Download

Share

Gmail Facebook Twitter LinkedIn More