Open AccessCOSM: Controlled Over-Sampling Method
Author(s) -
Gaetano Zazzaro
Publication year - 2020
Publication title -
transactions on machine learning and artificial intelligence
Language(s) - English
Resource type - Journals
ISSN - 2054-7390
DOI - 10.14738/tmlai.82.7925
Subject(s) - oversampling , computer science , class (philosophy) , data mining , sampling (signal processing) , artificial intelligence , machine learning , certainty , mathematics , bandwidth (computing) , computer network , geometry , filter (signal processing) , computer vision
The class imbalance problem is widespread in Data Mining and it can reduce the general performance of a classification model. Many techniques have been proposed in order to overcome it, thanks to which a model able to handling rare events can be trained. The methodology presented in this paper, called Controlled Over-Sampling Method (COSM), includes a controller model able to reject new synthetic elements for which there is no certainty of belonging to the minority class. It combines the common Machine Learning method for holdout with an oversampling algorithm, for example the classic SMOTE algorithm. The proposal explained and designed here represents a guideline for the application of oversampling algorithms, but also a brief overview on techniques for overcoming the problem of the unbalanced class in Data Mining.