Forecasting Graph Signals with Recursive MIMO Graph Filters
Signal Processing
2022-10-28 v1 Machine Learning
Abstract
Forecasting time series on graphs is a fundamental problem in graph signal processing. When each entity of the network carries a vector of values for each time stamp instead of a scalar one, existing approaches resort to the use of product graphs to combine this multidimensional information, at the expense of creating a larger graph. In this paper, we show the limitations of such approaches, and propose extensions to tackle them. Then, we propose a recursive multiple-input multiple-output graph filter which encompasses many already existing models in the literature while being more flexible. Numerical simulations on a real world data set show the effectiveness of the proposed models.
Cite
@article{arxiv.2210.15258,
title = {Forecasting Graph Signals with Recursive MIMO Graph Filters},
author = {Jelmer van der Hoeven and Alberto Natali and Geert Leus},
journal= {arXiv preprint arXiv:2210.15258},
year = {2022}
}