Kevin Slote kslote1

Kevin Slote 👋

Upcoming Talk: Attractor reconstruction and their topological data analysis for seizure onset detection

Biography

Dr. Kevin Slote is an applied mathematician and data scientist whose work combines nonlinear dynamical systems, causal discovery, topological data analysis, machine learning, and complex systems science. He earned a Ph.D. in Applied Mathematics from Georgia State University, where his doctoral research was advised by Prof. Igor Belykh. His research develops mathematical and computational methods for discovering governing equations, causal mechanisms, and predictive biomarkers from complex data.

Education Background

Applied Mathematics Ph.D. — 2025, Georgia State University Abstract Algebra M.S. — 2016, Georgia State University Mathematics and Statistics B.S., Minor in Physics — 2011, Georgia State University

Research Interests

Nonlinear dynamical systems
Causal discovery and causal inference
Data-driven dynamical systems
Topological data analysis
Generative AI and transformer models
Kolmogorov-Arnold Networks
Algebraic geometry, ergodic theory, and applied category theory

Research Overview

Currently, my research is at the Clarkson Center for Complex Systems Science, where I work on nonlinear dynamics, Kolmogorov-Arnold Networks, and open-source software for optimal causation entropy. My work includes work on Kolmogorov-Arnold Networks for Dynamics (KANDy), node-degree volatility for seizure-onset zone identification, topological biomarkers of epilepsy, and the causal effects of Twitter advocacy and media coverage on firearm acquisition in the United States. In addition to my academic research, I have extensive experience in the technology sector, having worked as a Principal Research Data Scientist on machine learning, natural language processing, transformer-based models, causal inference, and generative AI systems. I am a co-inventor on patents related to cybersecurity and AI-enabled security services. My research has received public and media attention for its use of causal network methods to examine the relationships among media narratives, online advocacy, and trends in firearm acquisition.

I developed and maintain the open-source Optimal Causation Entropy software and related methods for identifying causal structure in complex systems. This work supports broader efforts to develop interpretable tools for discovering directional influence in networks, oscillatory systems, and high-dimensional dynamical data. My experience in the technology industry has taught me the importance of teaching students how to write maintainable scientific software, as seen in the KANDy software, to prepare them for both scientific and industrial careers.