斯坦福课程:深度学习理论(附在线观看地址)

AI科技大本营 2018-05-16 16:45
关注 0

深度学习为什么能够在近几年取得如此大的进展?除了经验主义之外,我们还应该从理论层面去理解深度学习的发展。为此,斯坦福大学于去年秋天推出了一门名为 Theories of Deep Learning(深度学习理论) 的课程。


这门课程整体架构可以用下面这张图来表示:



课程在线观看:

  • Lecture 01 : Deep Learning Challenge. Is There Theory?


  • Lecture 02 : Overview of Deep Learning From a Practical Point of View


  • Lecture 03 : Harmonic Analysis of Deep Convolutional Neural Networks


  • Lecture 04 : Convnets from First Principles: Generative Models, Dynamic Programming & EM


  • Lecture 05 : When Can Deep Networks Avoid the Curse of Dimensionality and Other Theoretical Puzzles


  • Lecture 06 : Views of Deep Networksfrom Reproducing Kernel Hilbert Spaces


  • Lecture 07 : Understanding and Improving Deep Learning With Random Matrix Theory


  • Lecture 08:Topology and Geometry of Half-Rectified Network Optimization


  • Lecture 09 : What’s Missing from Deep Learning?


  • Lecture 10 : Convolutional Neural Networks in View of Sparse Coding


课程讲义地址:

https://stats385.github.io/lecture_slides

视频源地址:(感谢up主微博@爱可可-爱生活)

https://www.bilibili.com/video/av16136625/

推荐阅读书单和其他资源地址:

https://stats385.github.io/readings

文章被以下专辑收录
评论(1)
未登录
支持Markdown和数学公式,公式格式:\\(...\\)或\\[...\\]
谢泽标 2018-05-16 17:46