Learning to Reconstruct Missing Data from Spatiotemporal Graphs with Sparse Observations

Ivan Marisca, Andrea Cini, Cesare Alippi

Keywords: missing data, time series imputation, spatiotemporal graph neural networks

TL;DR: We propose a graph neural network that exploits a novel spatiotemporal attention mechanism to impute missing values leveraging only (sparse) valid observations.

Abstract: Modeling multivariate time series as temporal signals over a (possibly dynamic) graph is an effective representational framework that allows for developing models for time series analysis. Spatiotemporal graphs are often highly sparse, with time series characterized by multiple, concurrent, and even long sequences of missing data, e.g., due to the unreliable underlying sensor network. In this context, autoregressive models can be brittle and exhibit unstable learning dynamics. The objective of this paper is to tackle the problem of learning effective models to reconstruct, i.e., impute, missing data points by conditioning the reconstruction only on the available observations. In particular, we propose a novel class of attention-based architectures that, given a set of highly sparse discrete observations, learn a representation for points in time and space by exploiting a spatiotemporal propagation architecture aligned with the imputation task. Representations are learned end-to-end to reconstruct observations w.r.t. the corresponding sensor and its neighboring nodes. Compared to the state of the art, our model handles sparse data without propagating prediction errors or requiring a bidirectional model to encode forward and backward time dependencies. Empirical results on representative benchmarks show the effectiveness of the proposed method.

Type Of Submission: Extended abstract (max 4 main pages).

Agreement: Check this if you are okay with being contacted to participate in an anonymous survey.

Type Of Submission: Extended abstract.

Software: https://github.com/Graph-Machine-Learning-Group/spin

Community Implementations: [ 2 code implementations](https://www.catalyzex.com/paper/arxiv:2205.13479/code)