• Dnyaneshwari Kale
• Avishkar Mane
• Bhargav Deshmukh
• Chaitali Mahale
• H.R. Kulkarni, Nikita Bichitkar

Migraines affect a significant portion of the global population, with prevalence estimates ranging from 10%–18%, and impose a substantial burden on individuals and health systems. Traditional diagnostic methods rely largely on patient self-report and clinician judgment, which can lead to misclassification and delayed treatment. Advances in machine learning have shown promise in improving diagnostic accuracy, though existing models often suffer from trade-offs between performance and interpretability. This study proposes a hybrid model combining decision tree (DT) classifiers and neural networks (NN) for classifying migraine subtypes and healthy controls. The decision tree component provides transparent rules that can aid clinical insight, whereas the neural network captures non-linear, complex relationships in the data. A comprehensive dataset containing demographic, clinical symptoms, physiological signals, and patient history was utilized. After extensive preprocessing, feature selection, model training, and hyperparameter tuning, the hybrid model achieved accuracy of 92.1%, F1-score of 0.90, and AUC-ROC of 0.95, outperforming standalone decision tree and neural network baselines. The results underscore the potential of hybrid approaches in bridging the gap between performance and interpretability in medical classification tasks. Finally, we discuss clinical implications, limitations, and directions for future research.

migraine classification, decision tree, neural network, hybrid model, machine learning, interpretability, healthcare diagnostics

"A Hybrid Approach for Migraine Classification Using Decision Trees and Neural Networks", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 1, page no.89-96, January-2026, Available :http://www.jetir.org/papers/JETIRHG06010.pdf

"A Hybrid Approach for Migraine Classification Using Decision Trees and Neural Networks", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 1, page no. pp89-96, January-2026, Available at : http://www.jetir.org/papers/JETIRHG06010.pdf