Approximation theory · Neural networks
Sharp closure thresholds for two-hidden-layer ReLU networks
Sharp width thresholds for approximation by two-hidden-layer ReLU networks.
About
My research interests span geometry, cryptography, data mining, and machine learning.
My recent papers study discrete geometric inequalities, finite geometry, and the approximation properties of neural networks.
Papers & preprints
View allFinite geometry · Rank-metric codes
Completing the classification of maximum scattered linear sets in PG(1, q5)
The classification of maximum scattered linear sets in PG(1,q⁵).
Discrete geometry · Quantitative stability
Exact Coordinate-Balancing Calculus and Sharp Stability for Symmetric Lattice Cross-Covariograms
Coordinate balancing, equality certificates, and sharp stability for symmetric lattice cross-covariograms.