GALACTIC: Global and Local Agnostic Counterfactuals for Time-series Clustering

Unlock Time-Series Clarity: Actionable Counterfactuals for Enterprise AI

Time-series clustering is a fundamental tool for pattern discovery, yet existing explainability methods, primarily based on feature at-tribution or metadata, fail to identify the transitions that move an instance across cluster boundaries. While Counterfactual Explana-tions (CEs) identify the minimal temporal perturbations required to alter the prediction of a model, they have been mostly confined to supervised settings. This paper introduces GALACTIC, the first unified framework to bridge local and global counterfactual explain-ability for unsupervised time-series clustering. At instance level (local), GALACTIC generates perturbations via a cluster-aware op-timization objective that respects the target and underlying cluster assignments. At cluster level (global), to mitigate cognitive load and enhance interpretability, we formulate a representative CE selection problem. We propose a Minimum Description Length (MDL) objective to extract a non-redundant summary of global explanations that characterize the transitions between clusters. We prove that our MDL objective is supermodular, which allows the corresponding MDL reduction to be framed as a monotone sub-modular set function. This enables an efficient greedy selection algorithm with provable (1 - 1/e) approximation guarantees. Ex-tensive experimental evaluation on the UCR Archive demonstrates that GALACTIC produces significantly sparser local CEs and more concise global summaries than state-of-the-art baselines adapted for our problem, offering the first unified approach for interpreting clustered time-series through counterfactuals.