Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Hybrid Method of Selection Features to Improve Performance of Covid-19 Classification

Rektorat UDINUS 2022 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2022
Goals
Record
08b67dd0-28ff-490a-881c-6532f58ff86d
Published
2026-10-30

Universitas Dian Nuswantoro conducted a study to improve the performance of classification algorithms using a hybrid feature-selection method on Covid-19 data. The activity involved Rosidin, Sabir; Muljono; and Supriyanto, Catur as participants. The method combined SBS filtering and a wrapper technique. Its aim was to reduce the initial feature sub-space, described as N for the total number of features, to K for the selected features.

The hybrid process began with 8974 features. After the filtering process, 184 features remained. The wrapper technique was then applied and resulted in 170 selected features. Performance evaluation was subsequently conducted using SVM and KNN. Testing SVM with the big data produced an accuracy of 83.8%. Testing KNN with the parameter value K = 5 produced an accuracy of 79.5%. The classification of the K value was determined by the researcher.

The precision comparison reported 32.6% for KNN and 87.6% for SVM. Recall results were 14.2% for KNN and 20.1% for SVM. The F1-Score comparison reported 17.3% for KNN and 27.5% for SVM. These findings are the reported results of testing the hybrid method, which combined filtering and wrapper techniques, followed by performance evaluation of classification algorithms on Covid-19 data. The activity addresses the stated goal of improving classification performance through feature selection and testing with SVM and KNN.

Reference https://www.scopus.com/pages/publications/85141528299?origin=resultslist