RAG-LLM Powered Multi-Dataset Machine Learning Framework for Personalized Ayurvedic Migraine Management

DOI

10.2991/978-94-6239-791-0_13How to use a DOI?

Keywords

Migraine; Ayurveda; Retrieval Augmented Generation; Large Language Models; Machine Learning; Dosha Prediction

Abstract

Past several decades millions of people are suffering from various neurological disorders and migraine is one of the severe condition affecting overall population. Normally the medication includes allopathic drugs which provide temporary relief but unable to handle root cause and use of such drugs for longer time is life threatening. Ayurveda, historical roots from ancient Indian system has been proven to be better alternative medicine system in chronicle diseases. Tridosha analysis is the foundation behind Ayurvedic treatment which consists of determining imbalances in Dosha viz.Vata, Pitta and Kapha, Based on the Prakriti of human being personalized treatments are recommended which not only look into symptoms but also address the root causes of the problem. However, this treasure is not yet conceptualized in an appropriate intelligent digital health system. This work proposes a hybrid framework integrating Machine Learning (ML), Retrieval-Augmented Generation (RAG), and Large Language Models (LLMs) which provides personalized Ayurvedic recommendations to deal with migraine. Ensemble classifiers predicts Dosha imbalance based on patient symptoms and lifestyle features whereas RAG is extremely useful for retrieving relevant Ayurvedic therapies, herbal formulations and dietary guidance. Groq-powered LLaMA3 model uses this knowledge is for providing human readable recommendations. The proposed work has synthetic dataset of 5000 patients so it should be thought of as clinical diagnosis supporting tool not a substitute of diagnosis. It is an attempt to provide a research prototype which is perfect blend of traditional knowledge and contemporary technique demonstrating personalized, interpretable healthcare decisions.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Sakshi Hedke
AU  - Saniya Samant
AU  - Laxmi Bewoor
PY  - 2026
DA  - 2026/10/09
TI  - RAG-LLM Powered Multi-Dataset Machine Learning Framework for Personalized Ayurvedic Migraine Management
BT  - Proceedings of the International Conference on Artificial Intelligence and Machine Learning for Agro-Ecosystems (ICAIML-AE 2026)
PB  - Atlantis Press
SP  - 140
EP  - 156
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-791-0_13
DO  - 10.2991/978-94-6239-791-0_13
ID  - Hedke2026
ER  -