Dissertation defence (Computer Science): MSc Moein Khalighi
MSc Moein Khalighi defends the dissertation in Computer Science titled “Impact of Memory on Complex Dynamics: Case Studies in Ecology and Epidemiology” at the University of Turku on 28 November 2025 at 12.00 (University of Turku, Agora, XXI, Turku).
Opponent: Dr. Katharine Coyte (University of Manchester, United Kingdom)
Custos: Professor Leo Lahti (University of Turku)
Doctoral Dissertation at UTUPub: https://urn.fi/URN:ISBN:978-952-02-0345-0
***
Summary of the Doctoral Dissertation:
This dissertation asks: What changes when models do not forget the past? Most models assume only the present matters, but here the models let recent history weigh more than distant history, so the past fades yet still guides what happens next. The main findings show that in ecology, memory can shift how ecosystems respond to shocks by moving the thresholds for sudden change, altering stability and recovery, and changing how fast systems bounce back. In epidemics, memory in contact patterns, behavior, or the environment reshapes the rise and fall of disease, so outbreak size and timing can differ from standard forecasts.
The dissertation also presents a practical software tool for working with these memory-aware models and demonstrates their value on real data, including Ebola and COVID-19 case studies. The wider impact is direct. In public health, such models can support earlier and better targeted actions. In ecosystem management, they can help anticipate tipping points, choose safer margins, and plan steps that respect how past events still shape the present. The core message is clear across fields: history is not a detail, and when models remember, their answers change in ways that matter for science, policy, and everyday decisions.
Opponent: Dr. Katharine Coyte (University of Manchester, United Kingdom)
Custos: Professor Leo Lahti (University of Turku)
Doctoral Dissertation at UTUPub: https://urn.fi/URN:ISBN:978-952-02-0345-0
***
Summary of the Doctoral Dissertation:
This dissertation asks: What changes when models do not forget the past? Most models assume only the present matters, but here the models let recent history weigh more than distant history, so the past fades yet still guides what happens next. The main findings show that in ecology, memory can shift how ecosystems respond to shocks by moving the thresholds for sudden change, altering stability and recovery, and changing how fast systems bounce back. In epidemics, memory in contact patterns, behavior, or the environment reshapes the rise and fall of disease, so outbreak size and timing can differ from standard forecasts.
The dissertation also presents a practical software tool for working with these memory-aware models and demonstrates their value on real data, including Ebola and COVID-19 case studies. The wider impact is direct. In public health, such models can support earlier and better targeted actions. In ecosystem management, they can help anticipate tipping points, choose safer margins, and plan steps that respect how past events still shape the present. The core message is clear across fields: history is not a detail, and when models remember, their answers change in ways that matter for science, policy, and everyday decisions.
University Communications