Notes and Comments on S. Mallat’s Lectures at Collège de France (2018)
Résumé
The 2018 course by Stéphane Mallat, Professor at the Collège de France, focuses on the challenge of high-dimensional data, particularly in the face of the curse of dimensionality. It reviews Fourier Analysis, followed by Wavelet Analysis, and regression/classification tasks using kernels, before delving into neural networks. Gradient descent optimization is described along with its convergence properties. What sets this course apart is its mathematical perspective on 'Why does it work?' or 'What are the underlying information and structures for it to work?' This distinguishes it from courses on how to implement specific methods using certain libraries or code.
Origine : Fichiers produits par l'(les) auteur(s)