An Optimized Hybrid Approach for Feature Selection Based on Chi-Square and Particle Swarm Optimization Algorithms

Data, Год журнала: 2024, Номер 9(2), С. 20 - 20

Опубликована: Янв. 25, 2024

Feature selection is a significant issue in the machine learning process. Most datasets include features that are not needed for problem being studied. These irrelevant reduce both efficiency and accuracy of algorithm. It possible to think about feature as an optimization problem. Swarm intelligence algorithms promising techniques solving this This research paper presents hybrid approach tackling selection. A filter method (chi-square) two wrapper swarm (grey wolf (GWO) particle (PSO)) used different improve system execution time. The performance phases proposed assessed using distinct datasets. results show PSOGWO yields maximum boost 95.3%, while chi2-PSOGWO improvement 95.961% experimental performs better than compared approaches.

Язык: Английский

---

Applying Different Machine Learning Techniques for Prediction of COVID-19 Severity

IEEE Access, Год журнала: 2021, Номер 9, С. 135697 - 135707

Опубликована: Янв. 1, 2021

Due to the increase in number of patients who died as a result SARS-CoV-2 virus around world, researchers are working tirelessly find technological solutions help doctors their daily work. Fast and accurate Artificial Intelligence (AI) techniques needed assist decisions predict severity mortality risk patient. Early prediction patient would saving hospital resources decrease continual death by providing early medication actions. Currently, X-ray images used symptoms detecting COVID-19 patients. Therefore, this research, model has been built different levels risks for based on applying machine learning techniques. To build proposed model, CheXNet deep pre-trained hybrid handcrafted were applied extract features, two methods: Principal Component Analysis (PCA) Recursive Feature Elimination (RFE) integrated select most important then, six applied. For experiments proved that merging features have selected PCA RFE together (PCA + RFE) achieved best results with all classifiers compared using or individually. The XGBoost classifier performance merged where it accomplished 97% accuracy, 98% precision, 95% recall, 96% f1-score 100% roc-auc. Also, SVM carried out same some minor differences, but overall was good 99% On other hand, Extra Tree 99.6% measures.

Язык: Английский

---

Metaheuristics based COVID-19 detection using medical images: A review

Computers in Biology and Medicine, Год журнала: 2022, Номер 144, С. 105344 - 105344

Опубликована: Март 10, 2022

Язык: Английский

---

Feature selection of pre-trained shallow CNN using the QLESCA optimizer: COVID-19 detection as a case study

Applied Intelligence, Год журнала: 2023, Номер 53(15), С. 18630 - 18652

Опубликована: Фев. 6, 2023

Язык: Английский

---

COFE-Net: An ensemble strategy for Computer-Aided Detection for COVID-19

Measurement, Год журнала: 2021, Номер 187, С. 110289 - 110289

Опубликована: Окт. 15, 2021

Язык: Английский

---