Sinan Gunturk

Sinan Gunturk is affiliated with New York University, where he conducts research in applied mathematics and theoretical machine learning. His work focuses on approximation theory using polynomials with binary coefficients and its applications to neural networks.

His research interests include:

• Approximation Theory
• Quantized Neural Networks
• Bernstein Polynomials
• Theoretical Foundations of Deep Learning

His recent work explores how continuous functions can be approximated using polynomials with coefficients restricted to {+1, -1} in the Bernstein basis, and how such constructions can be used to realize one-bit neural networks. These contributions lie at the intersection of computational mathematics and efficient deep learning architectures.

He has published preprints on these topics in 2021, presenting both theoretical bounds and practical realizations. His research has implications for low-precision computing and energy-efficient AI.

There are no mentions of awards, students, or email in the available data.