Computational systems for modeling & dynamics
Building computational systems to model, measure, and understand complex dynamical processes.
cSYMd (Computational systems for modeling & dynamics) builds rigorous computational systems that model, measure, and interpret complex dynamical processes.
We work at the intersection of nonlinear dynamics, signal analysis, and reproducible software. Our research spans physiological and biomechanical signals, multi-source performance and health time-series, and computational environments that put advanced methods to work in real research workflows.
We apply the same dynamical and computational toolkit to human and multi-source time-series and to modern intelligent systems. We focus especially where modeling moves toward edge devices and people work in close partnership with machines.
The lab builds open methods, validates carefully, and ships tools others can inspect and extend. We contribute open-source software for dynamical systems analysis and welcome collaboration with students, researchers, and groups who model complex systems with rigor.
Principal investigator: Nathaniel T. Berry, PhD University of North Carolina at Greensboro (UNCG) School of Education; Information, Library, & Research Sciences College of Health and Human Sciences; Department of Kinesiology
We turn dynamical systems theory into usable computational pipelines. Our methods transfer across physiological data, edge-aware computation, and intelligent systems, and we design them to hold up under scrutiny.
We use recurrence quantification analysis (RQA/CRQA), entropy measures, and complexity assessment to expose temporal structure that linear summaries often miss, in natural signals and in high-dimensional computational processes.
We process and analyze biosignals such as PPG, ECG/HRV, gait, and related modalities. We extract quality-aware features and physiologically meaningful metrics that research teams can use with confidence.
We study athletic and health-related time-series by integrating training, wellness, and device data into coherent models of load, recovery, and readiness. When we add data-driven components, we keep them transparent and testable.
We build software kernels and analysis environments that make nonlinear methods portable, testable, and reusable. That work supports classical dynamical analysis and the study of learning systems and model-generated trajectories, from research notebooks to high-assurance stacks.
We put lab methods into practice through software and studies that stay open where possible and reproducible by design.
SymWorx is a modular open-source ecosystem (Rust core with Python bindings) for mathematical signal processing and nonlinear dynamics. It provides RQA/CRQA, peak detection, and interactive analysis tools that support reproducible research.
View on GitHubWe publish peer-reviewed work and pursue ongoing research in exercise physiology, biometrics, computational modeling, and nonlinear methods for human performance, health data, and related complex systems.
Google Scholar profileWe build frameworks and tools that study athletic load, readiness, and performance time-series with research clarity and practical measurement in training environments.
Active research · methods & softwareWe quantify stochasticity and structure in iterative language-model revision trajectories with nonlinear dynamical systems tools (RQA, entropy). We treat revision as a process we can measure, compare, and validate.
Experimental pipeline · model dynamics · open methodsSelected work only. Request full project details and collaboration notes as needed. We maintain open-source contributions independently under their own licenses.
Exploring graduate research, proposing a collaboration, contributing to open tools, or building methods that span physiological data and intelligent systems? Reach out.
Prefer email or a public profile?
cSYMd · RESEARCH LAB