BIO-642 / 1 credit

Teacher: Gerstner Wulfram

Language: English


Frequency

Only this year

Summary

The Loss Landscape of Neural Networks is in general non-convex and rough, but recent mathematical results lead provide insights of practical relevance. 9 online lectures, lecturers from NYU, Stanford, Shanghai, IST Austria, Google, Facebook, EPFL.

Content

Note

By the end of this course you should be able to explain recent results on the shape and convergence properties of the loss landscape in neural networks.

Keywords

Artificial Neural Networks, Deep Learning, Optimization, Gradient Descent, Loss landscape, Convergence, Convexity, Permutation Symmetry

Learning Prerequisites

Required courses

A Master-level class on Artificial Neural Networks or Machine Learning or Optimization

In the programs

  • Number of places: 30
  • Exam form: Term paper (session free)
  • Subject examined: State of the Art Topics in Neuroscience XIII
  • Lecture: 10 Hour(s)
  • Project: 15 Hour(s)

Reference week