Fractional-Order Compartmental Models: Stability, Optimal Control, and Neural Network Validation — with a Proposed Extension to Cancer Immunotherapy

Prelegent(ci)

Zakirullah

Afiliacja

School of Mathematical Sciences, University of Electronic Science and Technology of China

Język referatu

angielski

Termin

13 maja 2026 14:15

Link

https://uw-edu-pl.zoom.us/j/97187479863?pwd=4ERaSaNt4dbYv77P5T9Ru1AawSYnia.1

Informacje na temat wydarzenia

This is online meeting

Seminarium

Seminarium Zakładu Biomatematyki i Teorii Gier

---

In this talk, I present my recent work on mathematical modelling of infectious disease dynamics using fractional-order compartmental models with Caputo, Caputo–Fabrizio, and Atangana–Baleanu–Caputo operators. The models are analyzed for existence and uniqueness, positivity and boundedness, basic reproduction number R₀, and local and global stability via Lyapunov functions. Optimal control strategies are formulated using Pontryagin’s Maximum Principle and solved via the forward–backward sweep algorithm. For numerical validation, I use two complementary approaches: Physics-Informed Neural Networks (PINNs) for joint parameter identification, and Artificial Neural Networks (ANNs) for cross-validation of fractional solutions across multiple values of the fractional parameter. Results from models for Nipah virus, Chikungunya virus, Herpes Simplex Virus, and HIV will be presented. I conclude by outlining a proposed PhD research direction aimed at extending these mathematical tools to cancer immunotherapy models, building directly on the work of the Warsaw Biomathematics group

---