Zimeng Lyu

Assistant Professor

Kean University

I am an Assistant Professor in the Department of Computer Science and Technology at Kean University, where my research focuses on evolutionary neural architecture search for time-series forecasting, with an emphasis on ultra-lightweight models for edge deployment.

My work develops neuroevolution methods that design recurrent neural networks automatically, producing forecasting models that are compact enough to run on resource-constrained devices yet competitive with far larger architectures. I am also interested in online and non-stationary settings, where models must adapt as data distributions shift over time. This research has been applied to coal-fired power plant operations and to stock return prediction and portfolio trading, and my current work extends these ideas toward cross-disciplinary edge-AI applications.

At Kean, I teach Fundamentals of AI, graduate Data Mining, and software engineering, and I founded the Kean Cup AI Competition to give students across disciplines a venue for applied AI work. I also pursue research in cross-disciplinary AI education pedagogy, with the goal of broadening AI literacy beyond computer science majors. I earned my Ph.D. in Computer Science from the Rochester Institute of Technology.