Graph-based semi-supervised learning: A comprehensive review

Z Song, X Yang, Z Xu, I King - IEEE Transactions on Neural …, 2022 - ieeexplore.ieee.org
Semi-supervised learning (SSL) has tremendous value in practice due to the utilization of
both labeled and unlabelled data. An essential class of SSL methods, referred to as graph …

Contrastive and generative graph convolutional networks for graph-based semi-supervised learning

S Wan, S Pan, J Yang, C Gong - … of the AAAI conference on artificial …, 2021 - ojs.aaai.org
Abstract Graph-based Semi-Supervised Learning (SSL) aims to transfer the labels of a
handful of labeled data to the remaining massive unlabeled data via a graph. As one of the …

Poisson learning: Graph based semi-supervised learning at very low label rates

J Calder, B Cook, M Thorpe… - … Conference on Machine …, 2020 - proceedings.mlr.press
We propose a new framework, called Poisson learning, for graph based semi-supervised
learning at very low label rates. Poisson learning is motivated by the need to address the …

Improved spectral convergence rates for graph Laplacians on ε-graphs and k-NN graphs

J Calder, NG Trillos - Applied and Computational Harmonic Analysis, 2022 - Elsevier
In this paper we improve the spectral convergence rates for graph-based approximations of
weighted Laplace-Beltrami operators constructed from random data. We utilize regularity of …

Consistency of Lipschitz learning with infinite unlabeled data and finite labeled data

J Calder - SIAM Journal on Mathematics of Data Science, 2019 - SIAM
We study the consistency of Lipschitz learning on graphs in the limit of infinite unlabeled
data and finite labeled data. Previous work has conjectured that Lipschitz learning is well …

Graph-based active learning for semi-supervised classification of SAR data

K Miller, J Mauro, J Setiadi, X Baca… - Algorithms for …, 2022 - spiedigitallibrary.org
We present a novel method for classification of Synthetic Aperture Radar (SAR) data by
combining ideas from graph-based learning and neural network methods within an active …

Lipschitz regularity of graph Laplacians on random data clouds

J Calder, N García Trillos, M Lewicka - SIAM Journal on Mathematical Analysis, 2022 - SIAM
In this paper we study Lipschitz regularity of elliptic PDEs on geometric graphs, constructed
from random data points. The data points are sampled from a distribution supported on a …

Uniform convergence rates for Lipschitz learning on graphs

L Bungert, J Calder, T Roith - IMA Journal of Numerical Analysis, 2023 - academic.oup.com
Lipschitz learning is a graph-based semisupervised learning method where one extends
labels from a labeled to an unlabeled data set by solving the infinity Laplace equation on a …

Rates of convergence for Laplacian semi-supervised learning with low labeling rates

J Calder, D Slepčev, M Thorpe - Research in the Mathematical Sciences, 2023 - Springer
We investigate graph-based Laplacian semi-supervised learning at low labeling rates (ratios
of labeled to total number of data points) and establish a threshold for the learning to be well …

Hamilton-Jacobi equations on graphs with applications to semi-supervised learning and data depth

J Calder, M Ettehad - Journal of Machine Learning Research, 2022 - jmlr.org
Shortest path graph distances are widely used in data science and machine learning, since
they can approximate the underlying geodesic distance on the data manifold. However, the …