Login Search
FMDB Transactions on Sustainable Health Science Letters Submit Paper
    Crossref logo

    Archiving

    Portico

    Fernando Martins De Bulhão is a member of

    Crossref logo

Random Competition Wild Horse Optimizer and Improved Deep Belief Network Framework for Chronic Disease Detection

Authors:
R. Arulmathi, M. Sakthivanitha

Addresses:
Department of Computer Science, Vels Institute of Science, Technology and Advanced Studies, Chennai, Tamil Nadu, India. Department of Computer Applications, Vels Institute of Science, Technology and Advanced Studies, Chennai, Tamil Nadu, India.

Abstract:

The prevalence of Chronic Disease (CD) has increased worldwide, affecting all socioeconomic groups and extending to all regions. The World Health Organisation (WHO) states that early prevention is crucial for some CD, including diabetes mellitus, stroke, cancer, heart disease, kidney failure, and hypertension. Data quality and predictor selection, including machine learning, are key to CD prediction. Data quality is most affected by missing values, normalisation, feature selection, and imbalance. The purpose of CD prediction is to improve performance and address concerns in medical datasets. This research presents a new Deep Learning (DL) CD prediction framework. The Chronic Disease Progression Tracker Dataset is its source. In K-Nearest Neighbor (KNN)-based missing data imputation, missing values are replaced with the values of their k-nearest neighbors. Z-score normalization scales each dataset value to 0 and 1. RCWHO selects the optimal dataset features. Notable improvements in RCWHO include the Competition Waterhole Mechanism (CWM) and the Random Running Approach (RRA). CWM is suggested to improve the CD dataset's best-feature-selection exploitation behaviour. RRA balances exploration with exploitation. ISMOTE generates synthetic samples from the minority class to correct class imbalance in datasets. IDBN classifiers are advanced generative models with positive-etheric distributions for CD classification. Precision, recall, F1-score, accuracy, and error are used to compare classification algorithms in the Matrix Laboratory R2021a (MATLABR2021a) simulator.

Keywords: Chronic Disease Prediction; Missing Data Imputation; Data Normalisation; Feature Selection; Classification and Optimisation; Machine Learning; Deep Learning.

Received on: 27/04/2025, Revised on: 30/06/2025, Accepted on: 21/09/2025, Published on: 07/05/2026

DOI: 10.69888/FTSHSL.2026.000636

FMDB Transactions on Sustainable Health Science Letters, 2026 Vol. 4 No. 2, Pages: 102-118

  • Views : 130
  • Downloads : 11
Download PDF