|
Title:
|
EXTRACTION OF CLASSIFICATION RULES FROM IMBALANCED DATASETS USING GA-ILS |
|
Author(s):
|
Aouatef Mahani, Sadjia Benkhider and Ahmed Riadh Baba-Ali |
|
ISBN:
|
978-989-8533-69-2 |
|
Editors:
|
Pedro Isaías and Hans Weghorn |
|
Year:
|
2017 |
|
Edition:
|
Single |
|
Keywords:
|
Data Mining, Classification, Imbalanced Datasets, Memetic Algorithm, Iterated Local Search, Genetic Algorithm |
|
Type:
|
Short Paper |
|
First Page:
|
289 |
|
Last Page:
|
294 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
The classification is a task of Data Mining. It consists of the extraction of a set of rules from large datasets. However, some datasets are imbalanced and they contain majority instances and minority ones. It is worth noting that classical learning algorithms are known to have a bias towards majority class instances. If classification is applied to imbalanced datasets; it is called partial classification. Partial classification approaches are generally based on sampling methods or algorithmic methods. In this paper, we propose a new hybrid approach using a three-phase-rule-based extraction process which takes into account minority and majority instances in imbalanced binary datasets. Our approach has been tested on several imbalanced binary datasets with different values of imbalanced ratio. The obtained results show the efficiency of our approach compared to other results available in the literature. |
|
|
|
|
|
|
