Revisiting Multivariate Time Series Forecasting with Missing Values

2025/09/27 by Jie Yang, Y. Hu, Yang, Jie +9 · 1 citation
Computer Science · Decision Sciences · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Forecasting Techniques and Applications #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Stock Market Forecasting Methods #Time Series Analysis and Forecasting

paper · pdf · doi:10.48550/arxiv.2509.23494

openalex publication_date 2025/09/27 · openalex created_date 2025/10/19 · openalex updated_date 2026/07/28

Abstract

Missing values are common in real-world time series, and multivariate time series forecasting with missing values (MTSF-M) has become a crucial area of research for ensuring reliable predictions. To address the challenge of missing data, current approaches have developed an imputation-then-prediction framework that uses imputation modules to fill in missing values, followed by forecasting on the imputed data. However, this framework overlooks a critical issue: there is no ground truth for the missing values, making the imputation process susceptible to errors that can degrade prediction accuracy. In this paper, we conduct a systematic empirical study and reveal that imputation without direct supervision can corrupt the underlying data distribution and actively degrade prediction accuracy. To address this, we propose a paradigm shift that moves away from imputation and directly predicts from the partially observed time series. We introduce Consistency-Regularized Information Bottleneck (CRIB), a novel framework built on the Information Bottleneck principle. CRIB combines a unified-variate attention mechanism with a consistency regularization scheme to learn robust representations that filter out noise introduced by missing values while preserving essential predictive signals. Comprehensive experiments on four real-world datasets demonstrate the effectiveness of CRIB, which predicts accurately even under high missing rates. Our code is available in https://github.com/Muyiiiii/CRIB.

Citations

• Glocal Information Bottleneck for Time Series Imputation
• Bridging Past and Future: Distribution-Aware Alignment for Time Series Forecasting
• Cross-Domain Conditional Diffusion Models for Time Series Imputation
• R&D-Agent-Quant: A Multi-Agent Framework for Data-Centric Factors and Model Joint Optimization
• S4M: S4 for multivariate time series forecasting with Missing values
• FinTSB: A Comprehensive and Practical Benchmark for Financial Time Series Forecasting
• FinMamba: Market-Aware Graph Enhanced Multi-Level Mamba for Stock Movement Prediction
• TimeFilter: Patch-Specific Spatial-Temporal Graph Filtration for Time Series Forecasting
• WPMixer: Efficient Multi-Resolution Mixing for Long-Term Time Series Forecasting
• LSR-IGRU: Stock Trend Prediction Based on Long Short-Term Relationships and Improved GRU
• Are Language Models Actually Useful for Time Series Forecasting?
• Adaptive Multi-Scale Decomposition Framework for Time Series Forecasting
• GinAR: An End-To-End Multivariate Time Series Forecasting Model Suitable for Variable Missing
• A Unified Replay-based Continuous Learning Framework for Spatio-Temporal Prediction on Streaming Data
• TimeXer: Empowering Transformers for Time Series Forecasting with Exogenous Variables
• FourierGNN: Rethinking Multivariate Time Series Forecasting from a Pure Graph Perspective
• Missing Value Imputation for Multi-attribute Sensor Data Streams via Message Propagation (Extended Version)
• iTransformer: Inverted Transformers Are Effective for Time Series Forecasting
• Fully-Connected Spatial-Temporal Graph for Multivariate Time-Series Data
• Temporal Conditioning Spiking Latent Variable Models of the Neural Response to Natural Visual Scenes
• Graph Convolutional Networks for Traffic Forecasting with Missing Values
• A Time Series is Worth 64 Words: Long-term Forecasting with Transformers
• TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis
• Are Transformers Effective for Time Series Forecasting?
• Learning to Reconstruct Missing Data from Spatiotemporal Graphs with Sparse Observations
• SAITS: Self-attention-based imputation for time series
• Filling the G_ap_s: Multivariate Time Series Imputation by Graph Neural Networks
• CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputation
• Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting
• Interrupted time series analysis using autoregressive integrated moving average (ARIMA) models: a guide for evaluating large-scale health interventions
• Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
• Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
• Spectral Temporal Graph Neural Network for Multivariate Time-series Forecasting
• Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting
• Neural Controlled Differential Equations for Irregular Time Series
• Information bottleneck through variational glasses
• Joint Modeling of Local and Global Temporal Dynamics for Multivariate Time Series Forecasting with Missing Values
• GP-VAE: Deep Probabilistic Time Series Imputation
• BRITS: Bidirectional Recurrent Imputation for Time Series
• An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling
• Attention Is All You Need
• Modeling Long- and Short-Term Temporal Patterns with Deep Neural Networks
• Temporal Ensembling for Semi-Supervised Learning
• Relevant sparse codes with variational information bottleneck
• Deep Learning and the Information Bottleneck Principle
• Adam: A Method for Stochastic Optimization
• Learning with Pseudo-Ensembles
• The information bottleneck method

Cited by

• Glocal Information Bottleneck for Time Series Imputation
• Bridging Past and Future: Distribution-Aware Alignment for Time Series Forecasting

Related

• CoIFNet: A Unified Framework for Multivariate Time Series Forecasting with Missing Values
• S4M: S4 for multivariate time series forecasting with Missing values
• Is Precise Recovery Necessary? A Task-Oriented Imputation Approach for Time Series Forecasting on Variable Subset
• Merlin: Multi-View Representation Learning for Robust Multivariate Time Series Forecasting with Unfixed Missing Rates
• Probabilistic Imputation for Time-series Classification with Missing Data