Exploring Neural Granger Causality with xLSTMs: Unveiling Temporal Dependencies in Complex Data

2025-02-14

AI Summary

This work addresses the challenge of modeling long-range causal dependencies in nonlinear time series. We propose Granger Causal xLSTM (GC-xLSTM), the first framework integrating extended Long Short-Term Memory (xLSTM) networks with the Granger causality paradigm. To identify lagged causal variables adaptively, we introduce a dynamic Lasso-based sparse projection mechanism. Furthermore, we design an end-to-end jointly optimized architecture that preserves causal interpretability while enhancing robustness. Extensive experiments on three benchmark datasets demonstrate that GC-xLSTM significantly outperforms classical Granger methods and state-of-the-art RNN- and Transformer-based baselines. It accurately captures nonlinear, long-range causal relationships, achieving superior forecasting accuracy, strong interpretability of inferred causal structures, and improved stability under distributional shifts.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Causality in time series can be difficult to determine, especially in the presence of non-linear dependencies. The concept of Granger causality helps analyze potential relationships between variables, thereby offering a method to determine whether one time series can predict-Granger cause-future values of another. Although successful, Granger causal methods still struggle with capturing long-range relations between variables. To this end, we leverage the recently successful Extended Long Short-Term Memory (xLSTM) architecture and propose Granger causal xLSTMs (GC-xLSTM). It first enforces sparsity between the time series components by using a novel dynamic lass penalty on the initial projection. Specifically, we adaptively improve the model and identify sparsity candidates. Our joint optimization procedure then ensures that the Granger causal relations are recovered in a robust fashion. Our experimental evaluations on three datasets demonstrate the overall efficacy of our proposed GC-xLSTM model.

Problem

• Determining causality in time series with non-linear dependencies.
• Capturing long-range relations between variables in Granger causality.
• Improving model robustness and sparsity in identifying causal relations.

Innovation

• Extended Long Short-Term Memory (xLSTM) architecture
• Dynamic lasso penalty for sparsity enforcement
• Joint optimization for robust causality recovery