ProCTI: Prototype-Refined Global Conditioning for Diffusion-Based Time Series Imputation

2026-09-29

AI Summary

This study addresses the performance limitations of diffusion-based imputation when local observations are sparse or highly noisy, stemming from a lack of global structural information. To overcome this, we propose ProCTI, a framework that learns prototypes to retrieve global data priors and fuses them with local signals to enhance conditioning during the reverse diffusion process. Furthermore, ProCTI incorporates a prototype refinement mechanism and latent state modeling, accompanied by theoretical guarantees tailored to specific missing-data scenarios. By effectively compensating for insufficient local information through global conditioning, the proposed method significantly outperforms strong baselines under random missing settings while remaining competitive in attribute-missing scenarios. These results validate the critical role of global conditioning in improving imputation accuracy.

Abstract

Time series imputation has progressed from statistical and deep learning approaches to diffusion-based models, which have shown strong recent performance. Existing diffusion-based methods typically condition the reverse process using local contextual information from the current or neighbouring windows. Meanwhile, global dataset-level structure often remains implicit, limiting performance when local observations are sparse, noisy, or unrepresentative. To address this issue, we propose ProCTI, a diffusion-imputation framework that augments local conditioning with retrieved global dataset-level priors through learned prototypes. A hybrid conditioning mechanism integrates this global context with local signals during reverse diffusion, enabling more accurate reconstruction under varying missingness scenarios. Experiments across multiple benchmark datasets show that ProCTI outperforms strong baselines overall under random missingness, while remaining competitive under attribute-wise missingness. Furthermore, we use a latent-regime data model to characterise the precise conditions under which prototype-derived global conditioning provably improves imputation. We support this with a general theoretical analysis of local-global conditioning.

Problem

• Time Series Imputation
• Diffusion Models
• Global Conditioning
• Local Context

Innovation

• Diffusion-based imputation
• Prototype learning
• Global conditioning
• Time series imputation
• Hybrid conditioning mechanism