Data-driven algorithm design, reducing machine learning bias with truncated statistics, and the regularization effect of initial large learning rates—take a deep dive into these topics with Machine Learning Dept. at Carnegie Mellon University’s Nina Balcan, #MIT’s Costis Daskalakis, and #Stanford’s Tengyu Ma: via Microsoft Research
https://www.microsoft.com/en-us/research/video/ai-institute-geometry-of-deep-learning-2019-day-2-session-2/?OCID=msr_video_mlbias_aiinst_fb
https://www.microsoft.com/en-us/research/video/ai-institute-geometry-of-deep-learning-2019-day-2-session-2/?OCID=msr_video_mlbias_aiinst_fb
Microsoft Research
AI Institute "Geometry of Deep Learning" 2019 [Day 2 | Session 2] - Microsoft Research
Deep learning is transforming the field of artificial intelligence, yet it is lacking solid theoretical underpinnings. This state of affair significantly hinders further progress, as exemplified by time-consuming hyperparameters optimization, or the extraordinary…