Modelling Tinnitus Functional Index reduction using supervised machine learning algorithms

Authors

Edmund F. Agyemang, The University of Texas Rio Grande ValleyFollow

Document Type

Article

Publication Date

12-2024

Abstract

This study aims to model the reduction in the Tinnitus Functional Index (TFI) utilizing supervised machine learning algorithms, focusing primarily on Ordinary Least Squares (OLS), K-Nearest Neighbor (KNN), Ridge, and Lasso regressions. Our analysis highlighted Group, ISI, and SWLS as significant predictors of TFI reduction, identified through the best subset selection and confirmed by both forward and backward selection criteria in the OLS regression. Notably, the shrinkage methods, Ridge and Lasso regressions, demonstrated superior performance compared to OLS and KNN, with the Ridge regression presenting the smallest test mean square error (MSE) of 318.30. This finding establishes the Ridge regression as the best model for analyzing our Tinnitus dataset relative to the other methods, which exhibited test MSEs of 319.28 (Lasso), 330.76 (OLS), and 584.92 (KNN), respectively. This research highlights the potential of supervised machine learning algorithms in advancing personalized Tinnitus treatment, reflecting broader trends in the field as evidenced by studies in the literature. By leveraging these algorithms, we can enhance treatment precision and outcomes, contributing significantly to improved quality of life for individuals with Tinnitus. Future research should explore the integration of multimodal data and longitudinal applications of these algorithms to further refine predictive capabilities and treatment effectiveness.

Comments

© Edmund Fosu Agyemang. Article available under the CC BY-SA 4.0 licence

Recommended Citation

Agyemang, Edmund Fosu. "Modelling Tinnitus Functional Index reduction using supervised machine learning algorithms." Statistics in Transition. New Series 25, no. 4 (2024): 51-77. https://doi.org/10.59139/stattrans-2024-003

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 4.0 International License.

Publication Title

Statistics in Transition New Series

DOI

10.59139/stattrans-2024-003