Approximation Approaches - from Fourier Analysis to Deep Learning 231791-D
Wykład (WYK) Semestr letni 2022/23

Measure theory - Borel sets, Lebesgue measure and integral

Locally compact groups, Haar Theorem, Pontrjagin Theorem on Duality

Fourier Transformation - general approach and special cases

Fourier Transformation - properties, Bochner-Weil, Raykovwa and Herglotz Theorems.

Review of approximation theory - from orthogonal functions to neural networks

Convolution

Modern approximation Cybenko, Barron and Montufar Theorems

Principles of construction of deep learning models

Julia nad introduction Flux.jl library

Reinforcement learning, convolution and recurrent networks

Exercise - presentation of models (1)

Exercise - presentation of models (2)

Exercise - presentation of models (3)

Exercise - students presentations (1)

Exercise - students presentations (2)