An empowered AdaBoost algorithm implementation: A COVID-19 dataset study

•We propose a new AdaBoost algorithm for the classification of Covid-19 patients.•Decision Trees are integrated to boosting methodology.•The parameters of the algorithm are tuned after comprehensive experiments.•The proposed algorithm outperforms state-of-the-art classification algorithms.•The propo...

Full description

Saved in:
Bibliographic Details
Published in:Computers & industrial engineering Vol. 165; p. 107912
Main Author: Sevinç, Ender
Format: Journal Article
Language:English
Published: England Elsevier Ltd 01.03.2022
Subjects:
Adaptive boost algorithm
Artificial intelligence
Covid-19
Machine learning
Adaptive boost algorithm
Covid-19
Artificial intelligence
Machine learning
ISSN:0360-8352, 1879-0550, 1879-0550
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•We propose a new AdaBoost algorithm for the classification of Covid-19 patients.•Decision Trees are integrated to boosting methodology.•The parameters of the algorithm are tuned after comprehensive experiments.•The proposed algorithm outperforms state-of-the-art classification algorithms.•The proposed algorithm is shown to be a promising method for Covid-19 cases. The Covid-19 outbreak, which emerged in 2020, became the top priority of the world. The fight against this disease, which has caused millions of people’s deaths, is still ongoing, and it is expected that these studies will continue for years. In this study, we propose an improved learning model to predict the severity of the patients by exploiting a combination of machine learning techniques. The proposed model uses an adaptive boost algorithm with a decision tree estimator and a new parameter tuning process. The learning ratio of the new model is promising after many repeated experiments are performed by using different parameters to reduce the effect of selecting random parameters. The proposed algorithm is compared with other recent state-of-the-art algorithms on UCI data sets and a recent Covid-19 dataset. It is observed that competitive accuracy results are obtained, and we hope that this study unveils more usage of advanced machine learning approaches.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0360-8352
1879-0550
1879-0550
DOI:10.1016/j.cie.2021.107912