HELIX: Hybrid Encoding with Learnable Identity and Cross-dimensional Synthesis for Time Series Imputation

TL;DR

HELIX introduces a novel time series imputation method that uses learnable feature identities and hybrid attention to improve dependency modeling and achieve state-of-the-art results.

Contribution

It proposes a new approach with persistent feature embeddings and end-to-end learning of feature dependencies, outperforming existing methods.

Findings

HELIX surpasses 16 baselines on 5 datasets across 21 settings.

Learned feature identities align with physical and semantic structures.

HELIX effectively models arbitrary feature dependencies from temporal co-variation.

Abstract

Time series imputation benefits from leveraging cross-feature correlations, yet existing attention-based methods re-discover feature relationships at each layer, lacking persistent anchors to maintain consistent representations. To address this, we propose HELIX, which assigns each feature a learnable feature identity, a persistent embedding that captures intrinsic semantic properties throughout the network. Unlike graph-based methods that rely on predefined topology and assume homogeneous spatial relationships, HELIX learns arbitrary feature dependencies end-to-end from temporal co-variation, naturally handling datasets where features mix spatial locations with semantic variables. Integrated with hybrid temporal-feature attention, HELIX achieves the state-of-the-art performance, surpassing all 16 baselines on 5 public datasets across 21 experimental settings in our evaluation. Furthermore, our mechanistic analysis reveals that HELIX aligns learned feature identities and dependencies with latent physical and semantic structure progressively across layers, demonstrating that it more effectively translates cross-feature structure into imputation accuracy.