A Hybrid Intelligent System Based on Feature Selection and Ensemble Learning for Detecting Parkinson’s Disease

Articles in Press

Document Type : Research Article

Authors

1 Department of Computer Engineering, Islamic Azad University, Sari Branch, Sari, Iran

2 Department of Computer Engineering, Islamic Azad University, Babol Branch, Babol, Iran

Abstract

In recent years, Parkinson's disease (PD) has become a global health problem. Early diagnosis of the disease has a high impact on the quality of treatment. Various machine learning methods and classification algorithms have been proposed to enhance the accuracy in PD detection. Accordingly, this paper proposes a hybrid intelligent system, which involves preprocessing using normalization, feature selection using an Improved Binary Whale Optimization Algorithm (IBWOA), and classification using a New Ensemble Learning Strategy (NELS). In this paper, IBWOA was used to choose the optimal subset of features for prediction, while NELS was employed to handle the learning process. The PD dataset required for the purposes of this research was extracted from the UCI machine learning database. The experimental results showed that the combination of preprocessing, feature selection, and ensemble learning gave a classification accuracy of 96.9231% for the PD dataset. The results also showed that both the main phases of feature selection and ensemble learning are effective in improving system performance. The detection accuracy of the proposed system improved by 0.9231% compared to the best model in the current literature. 

Keywords

Main Subjects

Scientia Iranica

Articles in Press, Accepted Manuscript
Available Online from 11 May 2025

Files

History

  • Receive Date: 14 September 2024
  • Revise Date: 06 January 2025
  • Accept Date: 13 April 2025

Share

How to cite

Statistics