Parametric Feature Selection for an Enhanced Random Linear Oracle Ensemble Method
Keywords:
Ensemble, Feature Selection, Naïve Bayes, Pattern Recognition, Random Linear Oracle,Abstract
Random Linear Oracle (RLO) utilized classifier fusion-selection approach by replacing each classifier with two mini-ensembles separated by an oracle. This research investigates the effect of t-test feature selection toward classification performance of RLO ensemble method. Naïve Bayes (NB) classifier has been chosen as the base classifier due to its elegant simplicity and computationally inexpensive. Experiments were carried out using 30 data sets from UCI Machine Learning Repository. The results showed that RLO ensemble could greatly improve the ability of NB classifier in dealing with more data with different properties. Moreover, RLO ensemble receives benefits from feature selection algorithm, with a properly selected number of features from ttest, the performance of ensemble can be improved.Downloads
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2024 Journal of Telecommunication, Electronic and Computer Engineering
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)