Year: 2025
Type: article
Abstract: Background Dementia remains a global health challenge, with over 50 million affected individuals and projections suggesting a tripling by 2050. Early and accurate prediction is critical for effective management. The aim of this study is to explore advanced machine learning techniques to enhance predictive accuracy. Method The predictive capabilities of three machine learning models; Support Vector Regression (SVR), Gradient Boosting Machine (GBM), and Adaptive Neuro‐Fuzzy Inference System (ANFIS) in the context of dementia diagnosis were investigated. A dataset of 149 participants aged 60–96 years, incorporating nine clinical and imaging biomarkers, was utilized for model training and validation splits of 70% and 30%, respectively. Key evaluation metrics included R 2 , RMSE, and MSE. Result The results demonstrated that all models accurately predicted dementia, with ANFIS outperforming SVR and GBM in terms of precision and consistency. ANFIS achieved an R 2 of 1.0 during both training and testing phases, while GBM and SVR also displayed high accuracy with R 2 values of 0.998 and 0.995, respectively. The results validate the efficacy of hybrid and ensemble approaches like ANFIS for dementia prediction. Conclusion These findings underscore the potential of ensemble and hybrid machine learning approaches in enhancing dementia prediction accuracy, paving the way for early and reliable diagnosis.
Source: Alzheimer s & Dementia
Institutions: Near East University, University of Sharjah
Language: English
Subfield: Health Information Management
Field: Health Professions
Domain: Health Sciences
Open Access status: hybrid
Estimated APC paid: $4,200
OpenAlex ID: W7117253222
FWCI:
Cites:
Cited by:
Related to: