“MATH for Machine Learning” Seminar – On the Proof of Global Convergence of Gradient Descent for Deep ReLU Networks with Linear Widths Mr. Tianrui He (HKU) [LINK]