Part 3: VC Dimension - Neural Networks

By Mohammad Sadil Khan in Computational Theory Deep Learning VC Dimension MLP

September 30, 2022

This is the third part of the VC Dimension and Neural Networks series. This part discusses the important theorems for vc dimension of MLPs. Please read Part 2: VC Dimension - Theorems and Lemmas for the important theorems that will be used here.

1. Important Notation

AbbreviationMeaning
$NN$Neural Networks
$MLP$Multilayer Perceptrons

2. Theorem 1 - Separable Regions

In this theorem we will see how MLPs divides a space into different regions and classify accordingly.