Goldfeld's research interests include information theory, statistical machine learning, high-dimensional and nonparametric statistic, applied probability and interacting particle systems. Specifically, his recent work focuses on information-theoretic analysis of deep neural networks (DNNs), convergence rates of empirical measures smoothed by Gaussian kernels (which has to do with mutual information estimation between layers of a DNN), and data storage in stochastic Ising models. During his graduate studies, Goldfeld focused on physical layer security, cooperation in multiuser information-theoretic problems and multiuser channel and source duality.
