EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

From forward-backward sweep to backpropagation for optimal control in tumor-immune dynamics

Wenjing Zhang, Wandi Ding, Huaiping Zhu

Journal of Mathematical Biology · 2026-11 · 93 (5)

DOI: 10.1007/s00285-026-02475-2 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Mathematical Biology, online erschienen am 29.10.2026. Verfasst von Wenjing Zhang und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Mathematics and Statistics, Texas Tech University, Lubbock, TX, USA. wenjing.zhang@ttu.edu. Hauptthemen laut MEDLINE: Neoplasms, Models, Immunological. Die Arbeit nennt 56 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Simons Foundation, National Science Foundation, Natural Sciences and Engineering Research Council of Canada.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Optimal control problems in mathematical biology are commonly solved by the forward-backward sweep on the Pontryagin’s Maximum Principle necessary conditions. The sweep has known stability issues at constraint-saturated optima. We frame the same problem within the universal differential equation framework. A neural network represents the control. The cost functional serves as the training loss. We prove that any stationary point of the training loss satisfies the continuous first-order necessary conditions to a discretisation-controlled accuracy. On a grid of bifurcation-parameter and initial-condition scenarios drawn from a tumor-immune dynamics model, the standard sweep fails on most scenarios in the nominal regime. Neural network training reaches a stable optimum on every scenario, with a lower final cost than the converged damped sweep iterate. The pipeline accommodates cost functional variants by changing only the loss function. A Sobol sensitivity analysis confirms robustness of the trained schedule to perturbations of the cost weights.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

NeoplasmsModels, ImmunologicalMathematical ConceptsNeural Networks, ComputerHumansComputer SimulationAnimals

Stichwörter der Autor:innen

Forward-backward sweepPontryagin’s maximum principleScientific machine learningTumor-immune dynamicsUniversal differential equations

Autor:innen und Institutionen

  • Department of Mathematics and Statistics, Texas Tech University, Lubbock, TX, USA. wenjing.zhang@ttu.edu.
  • Department of Mathematical Sciences, Middle Tennessee State University, Murfreesboro, TN, USA.
  • Department of Mathematics and Statistics, York University, Toronto, ON, Canada.

Förderung

  • Simons Foundation
  • National Science Foundation
  • Natural Sciences and Engineering Research Council of Canada

Zitieren

APA

Zhang, W., Ding, W., & Zhu, H. (2026). From forward-backward sweep to backpropagation for optimal control in tumor-immune dynamics. Journal of Mathematical Biology, 93(5). https://doi.org/10.1007/s00285-026-02475-2

Vancouver

Zhang W, Ding W, Zhu H. From forward-backward sweep to backpropagation for optimal control in tumor-immune dynamics. Journal of Mathematical Biology. 2026;93(5). doi:10.1007/s00285-026-02475-2

BibTeX
@article{Zhang2026,
  author = {Zhang, Wenjing and Ding, Wandi and Zhu, Huaiping},
  title = {From forward-backward sweep to backpropagation for optimal control in tumor-immune dynamics},
  journal = {Journal of Mathematical Biology},
  year = {2026},
  volume = {93},
  number = {5},
  doi = {10.1007/s00285-026-02475-2}
}
Kostenlos starten →