Neutrosophic Cognitive Maps for Clinical Decision Making in Mental Healthcare: A Federated Learning Approach

Authors

  • Jagan M. Obbineni VIT School for Agricultural Innovations and Advanced Learning (VAIAL), Vellore Institute of Technology, Katpadi, Vellore, Tamil Nadu, 632014, India;
  • Ilanthenral Kandasamy School of Computer Science and Engineering (SCOPE), Vellore Institute of Technology, Katpadi, Vellore, Tamil Nadu, 632014, India;
  • Madhumitha Ramesh School of Computer Science and Engineering (SCOPE), Vellore Institute of Technology, Katpadi, Vellore, Tamil Nadu, 632014, India;
  • Florentin Smarandache Dept of Math and Sciences, University of New Mexico, Gallup, NM, United States;
  • Vasantha Kandasamy School of Computer Science and Engineering (SCOPE), Vellore Institute of Technology, Katpadi, Vellore, Tamil Nadu, 632014, India;

Keywords:

Federated Learning; Neutrosophic Cognitive Maps (NCMs); Mental Health; Psychological Concepts

Abstract

 Due to data privacy concerns and a lack of broadly applicable modelling approaches, 
mental health prediction encounters substantial challenges. This research introduces a pioneering 
decentralized framework integrating federated learning with Neutrosophic Cognitive Maps 
(NCMs) to facilitate secure and accurate mental health predictions while preserving data privacy. 
This innovative approach allows collaborative NCMs training on sensitive patient data across 
diverse sites without centralizing or transferring the data. The NCMs incorporated into the 
framework effectively model relationships between various symptoms and mental health states, 
offering interpretable insights into the complex dynamics of mental health. To address the 
limitations of local data availability, a multi-task learning methodology is employed, leveraging 
commonalities between related mental health prediction tasks to enhance modelling. Experiments 
are done on a synthetic mental health dataset to validate the proposed approach, demonstrating 
significant improvements. The decentralized nature of the approach ensures robust privacy 
guarantees by preventing direct access to patient data. The proposed framework contributes to the 
responsible application of soft computing and AI in the sensitive mental health domain. 
Furthermore, the interpretability of NCM models facilitates a nuanced analysis of indeterminate 
interrelationships between various psychological concepts, offering valuable support for 
data-driven decision-making in mental health contexts.

 

DOI: 10.5281/zenodo.10905866

Downloads

Download data is not yet available.

Downloads

Published

2024-04-02

How to Cite

Jagan M. Obbineni, Ilanthenral Kandasamy, Madhumitha Ramesh, Florentin Smarandache, & Vasantha Kandasamy. (2024). Neutrosophic Cognitive Maps for Clinical Decision Making in Mental Healthcare: A Federated Learning Approach . Neutrosophic Sets and Systems, 66, 1-11. https://fs.unm.edu/nss8/index.php/111/article/view/4357

Most read articles by the same author(s)

<< < 5 6 7 8 9 10 11 > >>