graph convolution

Data
Linear Shift-Invariant /Convolutional Graph Filters

We define our convolutional graph filter (or linear shift-invariant) as follows. Let S be a graph shift operator and x some graph signal. Then a filter H has

y=H(S)x=k=0K1hkSkx,h0,h1,,hK1R

Note that means H(S) is a (K1)th degree polynomial of S with (filter) coefficients hi
^definition

Properties

Mentions

File Last Edited
2025-03-04 equivariant lecture 6 2025-04-02
integral Lipschitz filters are stable to relative perturbations 2025-04-02
lipschitz graph filter 2025-04-14
the graph convolution converges to the graphon convolution in the spectral domain for a convergent sequence of graph signals 2025-04-14
ChebNet 2025-04-18
2025-01-27 graphs lecture 2 2025-05-13
2025-01-29 graphs lecture 3 2025-05-13
2025-02-03 graphs lecture 4 2025-05-13
2025-02-05 graphs lecture 5 2025-05-13
2025-02-12 graphs lecture 7 2025-05-13
2025-02-17 graphs lecture 8 2025-05-13
2025-02-19 graphs lecture 9 2025-05-13
2025-02-25 equivariant lecture 4 2025-05-13
2025-03-05 graphs lecture 12 2025-05-13
2025-03-10 graphs lecture 13 2025-05-13
2025-04-09 lecture 19 2025-05-13
conditions for finding a convolutional graph filter 2025-05-13
convolutional filter bank 2025-05-13
convolutional graph filters are local 2025-05-13
convolutional graph filters are permutation equivariant 2025-05-13
convolutional graph filters are shift equivariant 2025-05-13
discriminability of a graph filter 2025-05-13
filter permutation invariance 2025-05-13
GCN layers can be written as graph convolutions 2025-05-13
GNNs perform better than their constituent filters 2025-05-13
graph convolutions are stable to perturbations in the data and coefficients 2025-05-13
Graph Neural Networks 2025-05-13
graph perceptron 2025-05-13
graph SAGE 2025-05-13
graph signal processing problem 2025-05-13
Improved Image Classification with Manifold Neural Networks 2025-05-13
integral lipschitz filters are stable to dilations 2025-05-13
Lipschitz filters are stable to additive perturbations 2025-05-13
MPNNs can be expressed as graph convolutions 2025-05-13
node-level task 2025-05-13
quasi-symmetry 2025-05-13
spectral representation of a convolutional graph filter 2025-05-13
stable graph filter 2025-05-13
we can use GNNs to solve feature-aware semi-supervised learning problems 2025-05-13
We can verify whether graphs without node features and different laplacian eigenvalues are not isomorphic 2025-05-13
Graph Signals and Graph Signal Processing 2025-05-29
2025-03-24 graphs lecture 14 2025-05-30
2025-04-07 lecture 18 2025-05-30
convergence bound for graph convolutions 2025-05-30
graph convolutions of bandlimited signals converge to the graphon convolution 2025-05-30
fixed coefficients yield the same spectral response for both graphon and graph convolutions 2025-06-02
spectral representation of graphon convolutions 2025-06-02
lipschitz graph convolutions of graph signals converge to lipschitz graphon filters 2025-06-03
2025-03-26 lecture 15 2025-06-05
2025-04-02 lecture 17 2025-06-05