📚 Causality meets ST Data - Awesome Papers

Yutong Xia | Nov 23, 2024 min read

Awesome

[The Github repo can be found here!]

This is a curated collection of papers on the intersection of causality (including causal inference and causal discovery), spatio-temporal data (including spatio-temporal graph/series data, grid data, and trajectory data), and machine learning. For clarity, irregular grid data is categorized as spatio-temporal graph data here. In addition, some papers on multivariate time series, which share a similar data structure with spatio-temporal series, are also included.

Since the collection is curated by an individual, some important papers might have been unintentionally missed. Your contributions are greatly appreciated! Feel free to suggest additional papers through Issues or Pull Requests.

Survey & Tutorial

Survey

  • [2024] Causality for Earth Science – A Review on Time-series and Spatiotemporal Causality Methods [pdf] (Causal inference, Causal discovery, Spatio-temporal data)
  • [2023] Data-Driven Causal Effect Estimation Based on Graphical Causal Modelling: A Survey [pdf] (Causal inference)
  • [2023] Causal Discovery from Temporal Data: An Overview and New Perspectives [pdf] (Causal discovery, Time series)
  • [2023] A Survey on Causal Discovery Methods for I.I.D. and Time Series Data [pdf] (Causal discovery, Time series)
  • [2022] Causal Machine Learning: A Survey and Open Problems [pdf] (CausalML - causal supervised learning, causal generative modeling, causal explanations, causal fairness, causal reinforcement learning)
  • [2022] Survey and Evaluation of Causal Discovery Methods for Time Series [pdf] (Causal Discovery, Time Series)
  • [2021] D’ya like DAGs? A Survey on Structure Learning and Causal Discovery [pdf] (Causal discovery)

Tutorial

  • [SIGSPATIAL'24] Tutorial on Causal Inference with Spatiotemporal Data [pdf] [code]
  • [KDD'23] Causal Discovery from Temporal Data [pdf] [website]
  • [KDD'21] Causal Inference from Network Data [pdf]
  • [KDD'21] Causal Inference and Machine Learning in Practice with EconML and CausalML: Industrial Use Cases at Microsoft, TripAdvisor, Uber [pdf]

Papers

Causal Inference

Spatio-Temporal Graphs/Multivariate Time Series

  • [ICLR'24] Causality-Inspired Spatial-Temporal Explanations for Dynamic Graph Neural Networks [pdf] [code] (Node and graph classification, Back-door adjustment)
  • [AAAI'24] Instrumental Variable Estimation for Causal Inference in Longitudinal Data with Time-Dependent Latent Confounders [pdf] (Instrumental variable, Causal effect estimation)
  • [CIKM'24] Causality-Aware Spatiotemporal Graph Neural Networks for Spatiotemporal Time Series Imputation [pdf] (Imputation, Front-door adjustment, Causal attention)
  • [WSDM ’24] CityCAN: Causal Attention Network for Citywide Spatio-Temporal Forecasting [pdf] (Forecasting, Causal attention)
  • [BigData'24] Spatiotemporal Learning With Decoupled Causal Attention for Multivariate Time Series][pdf] (Forecasting, Causal attention)
  • [ICCBR'24] Spatio-Temporal Graph Neural Network with Hidden Confounders for Causal Forecast [pdf] [code] (Forecasting, Back-door criterion)
  • [arXiv'24] Causally-Aware Spatio-Temporal Multi-Graph Convolution Network for Accurate and Reliable Traffic Prediction [pdf] (Forecasting, Uncertainty quantification, Information fusion)
  • [NeurIPS'23] Deciphering Spatio-Temporal Graph Forecasting: A Causal Lens and Treatment [pdf] [code] (Forecasting, Back-door adjustment, Front-door adjustment)
  • [AAAI'23] Spatio-temporal neural structural causal models for bike flow prediction [pdf] [code] (Forecasting, Front-door criterion)
  • [AAAI'23] Causal conditional hidden Markov model for multimodal traffic prediction [pdf] [code] (Conditional Markov Process, Multimodal data)
  • [KDD'23] Generative Causal Interpretation Model for Spatio-Temporal Representation Learning [pdf] [code] (VAE, Forecasting, ICA Theory)
  • [ICMLA'23] Quantifying Causes of Arctic Amplification via Deep Learning based Time-series Causal Inference [pdf] [code] (Counterfactual prediction, Earth science)
  • [AAAG'23] Spatiotemporal Heterogeneities in the Causal Effects of Mobility Intervention Policies during the COVID-19 Outbreak: A Spatially Interrupted Time-Series (SITS) Analysis [pdf] (Spatio-temporal heterogeneity, Mobile phone data, Mobility control policy)
  • [Nature Reviews Earth & Environment’23] Causal inference for time series [pdf] [code] (Earth science, Causal effect estimation, Causal discovery)

Spatio-Temporal Grid

  • [NeurIPS'24] Causal Deciphering and Inpainting in Spatio-Temporal Dynamics via Diffusion Model [pdf] (Diffusion model, Backdoor adjustment, Frontdoor adjustment)
  • [ICLR'24] NuwaDynamics: Discovering and Updating in Causal Spatio-Temporal Modeling [pdf] [code] (Ocean system, Back-door adjustion)
  • [ECML'24] Estimating Direct and Indirect Causal Effects of Spatiotemporal Interventions in Presence of Spatial Interference [pdf] [code] (Treatment effects estimation, Spillover effects)
  • [ICML'22] CITRIS: Causal Identifiability from Temporal Intervened Sequences [pdf] [code] (Causal representations learning, Multidimensional causal factors, Image data)

Trajectory

  • [IJCAI'24] Towards Robust Trajectory Representations: Isolating Environmental Confounders with Causal Learning [pdf] (Representation learning, Intervention learning, Back-door adjustment)
  • [Inf.Fusion'24] Reliable trajectory prediction in scene fusion based on spatio-temporal Structure Causal Model [pdf] (Front-door criterion)

Causal Discovery

Spatio-Temporal Graphs/Multivariate Time Series

  • [AAAI'24] CUTS+: High-Dimensional Causal Discovery from Irregular Time-Series [pdf] [code] (Irregular sampling, Granger causality)
  • [ICLR'24] CausalTime: Realistically Generated Time-series for Benchmarking of Causal Discovery [pdf] [code] [websit] (Benchmarks, Time series)
  • [ICML'24] # Causal Discovery via Conditional Independence Testing with Proxy Variables [pdf] [code] (Time series, Proxy causal learning, Conditional independence testing)
  • [TNNLS ’24] Dynamic Causal Explanation Based Diffusion-Variational Graph Neural Network for Spatio-temporal Forecasting [pdf] [code] (Forecasting, Diffusion process)
  • [ICLR'23] CUTS: Neural Causal Discovery from Irregular Time-Series Data [pdf] [code] (EM-Style, Imputation, Irregular temporal data)
  • [NeurIPS'23] Causal Discovery from Subsampled Time Series with Proxy Variables [pdf] [code] (Time series, Subsampling)
  • [AAAI'23] Causal Recurrent Variational Autoencoder for Medical Time Series Generation [pdf] [code] (CR-VAE, Generative model)
  • [CIKM ’23] STREAMS: Towards Spatio-Temporal Causal Discovery with Reinforcement Learning for Streamflow Rate Prediction [pdf] [code] (Forecasting, Reinforcement Learning)
  • [CIKM ’23] Causal Discovery in Temporal Domain from Interventional Data [pdf][code] (Temporal reasoning, Multivariate time series)
  • [CIKM'22] Nonlinear Causal Discovery in Time Series [pdf] (Functional Causal Model, Time series, Non-stationary data)
  • [CLeaR'22] Amortized Causal Discovery: Learning to Infer Causal Graphs from Time-Series Data [pdf][code] (Granger Causality, Noisy Observations)
  • [ICLR'21] Interpretable Models for Granger Causality Using Self-explaining Neural Networks [pdf] [code] (Granger causality, Interpretable models)
  • [ICML'21] Necessary and sufficient conditions for causal feature selection in time series with latent common causes [pdf] (Causal feature selection)
  • [PAMI'21] Neural Granger Causality [pdf] [code] (Structured sparsity, Interpretability, Granger causality)
  • [NeurIPS'20] High-recall causal discovery for autocorrelated time series with latent confounders [pdf] [code] (Tigramite - Benchmark and python package, High-dimensional time series)
  • [ICML'20] CAUSE: Learning Granger Causality from Event Sequences using Attribution Methods [pdf] [code] (Granger causality)
  • [AISTATS'20] DYNOTEARS: Structure Learning from Time-Series Data [pdf] (Dynamic Bayesian networks, Structure learning, Time series, Acyclicity constraint)
  • [UAI'20] Discovering Contemporaneous and Lagged Causal Relations in Autocorrelated Nonlinear Time Series Datasets [pdf] [code] (Time series, Autocorrelation)
  • [Chaos'18] Causal network reconstruction from time series: From theoretical assumptions to practical estimation [pdf] [code] (Granger causality, Causal Markov condition)
  • [BigData'17] pg-Causality: Identifying Spatiotemporal Causal Pathways for Air Pollutants with Urban Big Data [pdf] (Bayesian learning, Causal pathways, Air Quality)

Spatio-Temporal Grid

  • [arXiv'24] Discovering Latent Structural Causal Models from Spatio-Temporal Data [pdf] [code] (Climate data)
  • [BigData'22] A spatio-temporal causal discovery framework for hydrological systems [pdf] [code] (Hydrological systems)
  • [Environmental Data Science'22] A spatiotemporal stochastic climate model for benchmarking causal discovery methods for teleconnections [pdf] [code] (Climate data, Teleconnections)
  • [Nature Communications'15] Identifying causal gateways and mediators in complex spatio-temporal systems [pdf] (Atmospheric dynamics, Complex systems, Causal effect estimation)