Transforming Causality: Transformer-Based Temporal Causal Discovery with Prior Knowledge Integration

• URL: http://arxiv.org/abs/2508.15928v1
• Date: Thu, 21 Aug 2025 19:19:11 GMT
• Title: Transforming Causality: Transformer-Based Temporal Causal Discovery with Prior Knowledge Integration
• Authors: Jihua Huang, Yi Yao, Ajay Divakaran,
• Abstract summary: We introduce a novel framework for temporal causal discovery and inference.<n>It addresses two key challenges: complex nonlinear dependencies and spurious correlations.<n>Our method significantly outperforms other state-of-the-art approaches.
• Score: 8.412444798554143
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We introduce a novel framework for temporal causal discovery and inference that addresses two key challenges: complex nonlinear dependencies and spurious correlations. Our approach employs a multi-layer Transformer-based time-series forecaster to capture long-range, nonlinear temporal relationships among variables. After training, we extract the underlying causal structure and associated time lags from the forecaster using gradient-based analysis, enabling the construction of a causal graph. To mitigate the impact of spurious causal relationships, we introduce a prior knowledge integration mechanism based on attention masking, which consistently enforces user-excluded causal links across multiple Transformer layers. Extensive experiments show that our method significantly outperforms other state-of-the-art approaches, achieving a 12.8% improvement in F1-score for causal discovery and 98.9% accuracy in estimating causal lags.

Related papers

• Transformer Is Inherently a Causal Learner [27.79148022495734]
  We show that transformer trained in an autoregressive manner naturally encodes time-delayed causal structures.<n>We prove this connection theoretically under standard identifiability conditions.<n>This approach greatly surpasses the performance of state-of-the-art discovery algorithms.
  arXiv  Detail & Related papers  (2026-01-09T09:10:04Z)
• A Reverse Causal Framework to Mitigate Spurious Correlations for Debiasing Scene Graph Generation [59.473751744275496]
  Scene Graph Generation (SGG) frameworks typically incorporate a detector to extract relationship features and a classifier to categorize these relationships.<n>Such a causal chain structure can yield spurious correlations between the detector's inputs and the final predictions.<n>We propose reconstructing the causal chain structure into a reverse causal structure, wherein the classifier's inputs are treated as the confounder.
  arXiv  Detail & Related papers  (2025-05-29T13:57:01Z)
• Transformers with Sparse Attention for Granger Causality [0.8249694498830561]
  Deep learning based methods such as transformers are increasingly used to capture temporal dynamics and causal relationships beyond mere correlations. Recent works suggest self-attention weights of transformers as a useful indicator of causal links. We propose a novel modification to the self-attention module to establish causal links between the variables of time-series data with varying lag dependencies.
  arXiv  Detail & Related papers  (2024-11-20T12:34:06Z)
• Precipitation Nowcasting Using Diffusion Transformer with Causal Attention [3.9468501770612576]
  Current deep learning methods fall short in establishing effective dependencies between conditions and forecast results. We propose a Precipitationcasting Using Diffusion Transformer with Causal Attention model.
  arXiv  Detail & Related papers  (2024-10-17T08:10:41Z)
• CausalFormer: An Interpretable Transformer for Temporal Causal Discovery [24.608536564444137]
  We propose an interpretable transformer-based causal discovery model termed CausalFormer. The causality-aware transformer learns the causal representation of time series data using a prediction task. The decomposition-based causality detector interprets the global structure of the trained causality-aware transformer.
  arXiv  Detail & Related papers  (2024-06-24T15:09:29Z)
• On the Identification of Temporally Causal Representation with Instantaneous Dependence [50.14432597910128]
  Temporally causal representation learning aims to identify the latent causal process from time series observations. Most methods require the assumption that the latent causal processes do not have instantaneous relations. We propose an textbfIDentification framework for instantanetextbfOus textbfLatent dynamics.
  arXiv  Detail & Related papers  (2024-05-24T08:08:05Z)
• Multi-modal Causal Structure Learning and Root Cause Analysis [67.67578590390907]
  We propose Mulan, a unified multi-modal causal structure learning method for root cause localization. We leverage a log-tailored language model to facilitate log representation learning, converting log sequences into time-series data. We also introduce a novel key performance indicator-aware attention mechanism for assessing modality reliability and co-learning a final causal graph.
  arXiv  Detail & Related papers  (2024-02-04T05:50:38Z)
• Correlation-aware Spatial-Temporal Graph Learning for Multivariate Time-series Anomaly Detection [67.60791405198063]
  We propose a correlation-aware spatial-temporal graph learning (termed CST-GL) for time series anomaly detection. CST-GL explicitly captures the pairwise correlations via a multivariate time series correlation learning module. A novel anomaly scoring component is further integrated into CST-GL to estimate the degree of an anomaly in a purely unsupervised manner.
  arXiv  Detail & Related papers  (2023-07-17T11:04:27Z)
• Causal discovery for time series with constraint-based model and PMIME measure [0.0]
  We present a novel approach for discovering causality in time series data that combines a causal discovery algorithm with an information theoretic-based measure. We evaluate the performance of our approach on several simulated data sets, showing promising results.
  arXiv  Detail & Related papers  (2023-05-31T09:38:50Z)
• DOMINO: Visual Causal Reasoning with Time-Dependent Phenomena [59.291745595756346]
  We propose a set of visual analytics methods that allow humans to participate in the discovery of causal relations associated with windows of time delay. Specifically, we leverage a well-established method, logic-based causality, to enable analysts to test the significance of potential causes. Since an effect can be a cause of other effects, we allow users to aggregate different temporal cause-effect relations found with our method into a visual flow diagram.
  arXiv  Detail & Related papers  (2023-03-12T03:40:21Z)
• Valid Inference After Causal Discovery [73.87055989355737]
  We develop tools for valid post-causal-discovery inference. We show that a naive combination of causal discovery and subsequent inference algorithms leads to highly inflated miscoverage rates.
  arXiv  Detail & Related papers  (2022-08-11T17:40:45Z)
• Path Signature Area-Based Causal Discovery in Coupled Time Series [0.0]
  We propose the application of confidence sequences to analyze the significance of the magnitude of the signed area between two variables. This approach provides a new way to define the confidence of a causal link existing between two time series.
  arXiv  Detail & Related papers  (2021-10-23T19:57:22Z)
• Transformer Hawkes Process [79.16290557505211]
  We propose a Transformer Hawkes Process (THP) model, which leverages the self-attention mechanism to capture long-term dependencies. THP outperforms existing models in terms of both likelihood and event prediction accuracy by a notable margin. We provide a concrete example, where THP achieves improved prediction performance for learning multiple point processes when incorporating their relational information.
  arXiv  Detail & Related papers  (2020-02-21T13:48:13Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.