FADTI: Fourier and Attention Driven Diffusion for Multivariate Time Series Imputation

TL;DR

FADTI introduces a diffusion-based multivariate time series imputation method that incorporates frequency-aware features via a learnable Fourier Bias Projection, enhancing performance especially under high missing data scenarios.

Contribution

The paper presents FADTI, a novel framework combining frequency-informed feature modulation with diffusion and attention mechanisms for improved time series imputation.

Findings

FADTI outperforms state-of-the-art methods on multiple benchmarks.

FADTI performs particularly well under high missing data rates.

The method effectively models both stationary and non-stationary patterns.

Abstract

Multivariate time series imputation is fundamental in applications such as healthcare, traffic forecasting, and biological modeling, where sensor failures and irregular sampling lead to pervasive missing values. However, existing Transformer- and diffusion-based models lack explicit inductive biases and frequency awareness, limiting their generalization under structured missing patterns and distribution shifts. We propose FADTI, a diffusion-based framework that injects frequency-informed feature modulation via a learnable Fourier Bias Projection (FBP) module and combines it with temporal modeling through self-attention and gated convolution. FBP supports multiple spectral bases, enabling adaptive encoding of both stationary and non-stationary patterns. This design injects frequency-domain inductive bias into the generative imputation process. Experiments on multiple benchmarks, including a newly introduced biological time series dataset, show that FADTI consistently outperforms state-of-the-art methods, particularly under high missing rates. Code is available at https://anonymous.4open.science/r/TimeSeriesImputation-52BF