معرفی
Jonathan Siegel is Assistant Professor of Mathematics at Texas A&M University. His research bridges approximation theory, neural network mathematics, statistics, and numerical methods for PDEs, with applications to materials science. Funded by NSF and ONR grants, his work develops theoretical foundations for deep learning algorithms.
Research areas include:
- Mathematical theory of neural networks
- High-order approximation rates for shallow networks
- Sparse neural network training
- Optimization on manifolds
- Structure-informed materials prediction
Recent publications analyze spectral bias of neural networks, approximation rates for ReLU networks, and greedy training algorithms.
۰مقاله منتشرشده
