
Zimeng Lyu
Ph.D. Candidate
Ph.D. Candidate in Computer Science specializing in time series forecasting, AutoML, Neural Architecture Search, and Transformers. Delivered $7.3M in power plant cost savings and beat stock market index with evolved RNNs. Proficient in C++, Python, PyTorch, Docker, Spark, and distributed computing for large-scale AI solutions. Serve as course designer and instructor for a graduate core course and graduate student research mentor.
Thesis Advisor: Dr. Travis Desell.
Email: zimenglyu@mail.rit.edu
Research Areas
Time Series Forecasting
Online and Offline Time Series Forecasting, Stock Return Predictions
Neural Architecture Search
NeuroEvolution, Recurrent Neural Networks
Minimally Supervised Learning
Self Organizing Maps, Topological Projections