Скачать с ютуб JORGE NOCEDAL | Optimization methods for TRAINING DEEP NEURAL NETWORKS в хорошем качестве

JORGE NOCEDAL | Optimization methods for TRAINING DEEP NEURAL NETWORKS

Conferencia "Optimization methods for training deep neural networks", impartida por el Dr. Jorge Nocedal (McCormick School of Engineering at Northwestern University) el 3 de diciembre de 2018, en el marco del 20 aniversario de la Maestría de Ciencias de la Computación del CIMAT. Resume: Most high-dimensional nonconvex optimization problems cannot be solved to optimality. However, deep neural networks have a benign geometry that allows stochastic optimization methods find acceptable solutions. There are, nevertheless, many open questions concerning the optimization process, including trade-offs between parallelism and the predictive ability of solutions, as well as the choice of a metric with the right statistical properties. In this talk we discuss classical and new optimization methods in the light of these observations, and conclude with some open questions.