MS04 - MEPI-08 Part 2 of 3

Modeling Complex Adaptive Systems in Life and Social Sciences (Part 2)

Tuesday, July 15 at 4:00pm

SMB2025 Follow

Note: this minisymposia has multiple sessions. The other sessions are: Part-1 and Part-3.

Yun Kang (Arizona State University), Tao Feng, Yangzhou University & University of Alberta

Description:

Utilizing complex adaptive systems in modeling has proven to be a powerful approach for understanding various aspects of life and social sciences across spatial and temporal scales. This special session will bring together a distinguished and diverse group of scholars from mathematics, biology, ecology, and epidemiology. These experts apply mathematical models and theoretical analysis to gain insights into critical biological, epidemiological, and social challenges. The session aims to provide an effective platform for presenting and discussing the latest research, fostering collaboration among professionals from different universities and career stages. Our goal is to encourage a rich exchange of ideas by assembling a group of researchers with diverse backgrounds, with a particular emphasis on promoting minority representation. The invited speakers span institutions across multiple countries and include individuals at various career stages, from early-career researchers to senior scholars. This inclusive approach ensures equal opportunities for all participants to present their findings and engage in meaningful collaborations.

Yaqi Chen

Harbin Institute of Technology & University of Alberta

"Well-Posedness and Dynamical Behavior of a Two-Species Reaction-Diffusion Model with Nonlocal Perception"

Nonlocal cues, such as visual, auditory, olfactory, and chemosensory cues, play a vital role in informing animal movement. To characterize these ecological phenomena, we propose a two-species reaction-diffusion model with nonlocal perception in a two-dimensional square domain. In this talk, I will first discuss the well-posedness of the proposed model, which is established using the entropy method. Subsequently, taking the predator-prey system as an illustrative example, I will conduct linear stability and bifurcation analyses by selecting the perception diffusion coefficient as a bifurcation parameter. The conditions for stability of positive constant steady states, as well as for the existence of Turing instability and Turing-Hopf bifurcations, will be identified explicitly. Furthermore, numerical simulations of the predator-prey model with a Holling type II functional response will be presented, including spatially nonhomogeneous steady-state patterns and spatially nonhomogeneous periodic patterns. These results highlight a complementary mechanism between perceptual range and perception diffusion ability, offering new theoretical insights and quantitative understanding of the role of nonlocal perception in spatial self-organization and pattern formation in ecological systems. This is joint work with Ben Niu and Hao Wang.

Shan Gao

University of Alberta

"Early detection of disease outbreaks and non-outbreaks using incidence data: A framework using feature-based time series classification and machine learning"

Forecasting the occurrence and absence of novel disease outbreaks is essential for disease management, yet existing methods are often context-specific, require a long preparation time, and non-outbreak prediction remains understudied. To address this gap, we propose a novel framework using a feature-based time series classification (TSC) method to forecast outbreaks and non-outbreaks. We tested our methods on synthetic data from a Susceptible–Infected–Recovered (SIR) model for slowly changing, noisy disease dynamics. Outbreak sequences give a transcritical bifurcation within a specified future time window, whereas non-outbreak (null bifurcation) sequences do not. We identified incipient differences, reflected in 22 statistical features and 5 early warning signal indicators, in time series of infectives leading to future outbreaks and non-outbreaks. Classifier performance, given by the area under the receiver-operating curve (AUC), ranged from 0.99 for large expanding windows of training data to 0.7 for small rolling windows. The framework is further evaluated on four empirical datasets: COVID-19 incidence data from Singapore, 18 other countries, and Edmonton, Canada, as well as SARS data from Hong Kong, with two classifiers exhibiting consistently high accuracy. Our results highlight detectable statistical features distinguishing outbreak and non-outbreak sequences well before potential occurrence, in both synthetic and real-world datasets presented in this study.

Bo-Wei Qin

Fudan University

"Polarization Does Not Necessarily Imply Conflict: Modeling and Modulating Pattern Boundaries of Opinion Dynamics"

Divergent opinions resulting from polarization are widespread across various fields, including economics, technology, and politics, and are often considered as social threats. Numerous studies were therefore devoted to achieving consensus. However, as we will discuss in this talk, polarization is inevitable when individuals exhibit black-and-white thinking (BWT), a previously underappreciated mechanism that drives adaptive consolidation of opinions. We will also demonstrate that the primary social threats do not arise directly from polarization itself, but rather from conflicts between connected individuals holding divergent opinions. By developing a networked dynamical model incorporating BWT and analyzing the pattern boundaries, we find that polarization does not necessarily imply conflict. Instead, the conflict intensifies through accumulating unstable eigenmodes, a process that is greatly influenced by network topology. This finding helps us elucidate how conflicts evolve across different social networks, and, more importantly, provides insights into developing effective modulation strategies to mitigate conflicts, even when polarization persists.

Joan Ponce

Arizona State University

"Extreme geographic misalignment of healthcare resources and HIV treatment deserts in Malawi"

The Joint United Nations Programme on HIV and AIDS has proposed that human rights should be at the center of efforts to end the HIV pandemic and achieving equity in access to antiretroviral therapy (ART) and HIV healthcare is essential. Here we present a geospatial and geostatistical modeling framework for conducting, at the national level, an equity evaluation of access to ART. We apply our framework to Malawi, where HIV prevalence is ~9%. Access depends upon the number of available healthcare facilities (HCFs), the travel times needed to reach these HCFs, the mode of transportation used (walking, biking, driving) and the supply-to-demand ratio for ART at the HCFs. We find extreme inequities in access to ART. Access maps show striking geographic patterns, revealing clusters of communities with very low or high levels of access. We discover that an extreme geographic misalignment of healthcare resources with respect to need has generated a new type of medical desert: an HIV treatment desert. Around 23% of people living with HIV reside in deserts where they have to walk up to 3 h to reach HCFs; in 2020, these HCFs only received 3% of the national supply of ART. We recommend strategies for shrinking deserts; if not implemented, deserts will grow in size and number.

#SMB2025 Follow

Annual Meeting for the Society for Mathematical Biology, 2025.