# Unsupervised Deep Haar Scattering on Graphs

arXiv: https://arxiv.org/abs/1406.2390 · HTML: https://mlsanity.com/paper/1406.2390/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 9 Jun 2014.

## Built on by (most cited first, 6 of 6)

- [Multiresolution Equivariant Graph Variational Autoencoder](https://mlsanity.com/paper/2106.00967/) (2021)
- [Do We Need an Encoder-Decoder to Model Dynamical Systems on Networks?](https://mlsanity.com/paper/2305.12185/) (2023)
- [Link Prediction without Graph Neural Networks](https://mlsanity.com/paper/2305.13656/) (2023)
- [Hierarchical Invariance for Robust and Interpretable Vision Tasks at Larger Scales](https://mlsanity.com/paper/2402.15430/) (2024)
- [Attribute-Enhanced Similarity Ranking for Sparse Link Prediction](https://mlsanity.com/paper/2412.00261/) (2024)
- [MIDI-RAE-JEPA: Hierarchical Representation Learning and Generation for Symbolic Music](https://mlsanity.com/paper/2607.14537/) (2026)

## Abstract (arXiv)

The classification of high-dimensional data defined on graphs is particularly difficult when the graph geometry is unknown. We introduce a Haar scattering transform on graphs, which computes invariant signal descriptors. It is implemented with a deep cascade of additions, subtractions and absolute values, which iteratively compute orthogonal Haar wavelet transforms. Multiscale neighborhoods of unknown graphs are estimated by minimizing an average total variation, with a pair matching algorithm of polynomial complexity. Supervised classification with dimension reduction is tested on data bases of scrambled images, and for signals sampled on unknown irregular grids on a sphere.