IPatch: A Multi-Resolution Transformer Architecture for Robust Time-Series Forecasting

2026-03-25

AI Summary

Multivariate time series forecasting faces significant challenges in simultaneously capturing short-term fluctuations and long-term dependencies, as well as balancing accuracy, efficiency, and robustness. This work proposes a novel multi-resolution Transformer architecture that uniquely integrates point-wise and patch-wise embeddings to construct a multi-granular temporal representation. By synergistically combining fine-grained detail preservation with enhanced contextual modeling, the proposed approach improves both predictive performance and computational efficiency. Extensive experiments demonstrate that the method consistently outperforms single-representation baselines across seven benchmark datasets, achieving notable gains in forecasting accuracy, noise robustness, and generalization capability across varying prediction horizons.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Computer Vision: Multi-modal Vision
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Accurate forecasting of multivariate time series remains challenging due to the need to capture both short-term fluctuations and long-range temporal dependencies. Transformer-based models have emerged as a powerful approach, but their performance depends critically on the representation of temporal data. Traditional point-wise representations preserve individual time-step information, enabling fine-grained modeling, yet they tend to be computationally expensive and less effective at modeling broader contextual dependencies, limiting their scalability to long sequences. Patch-wise representations aggregate consecutive steps into compact tokens to improve efficiency and model local temporal dynamics, but they often discard fine-grained temporal details that are critical for accurate predictions in volatile or complex time series. We propose IPatch, a multi-resolution Transformer architecture that integrates both point-wise and patch-wise tokens, modeling temporal information at multiple resolutions. Experiments on 7 benchmark datasets demonstrate that IPatch consistently improves forecasting accuracy, robustness to noise, and generalization across various prediction horizons compared to single-representation baselines.

Problem

• time-series forecasting
• temporal dependencies
• multi-resolution representation
• Transformer architecture
• multivariate time series

Innovation

• multi-resolution
• time-series forecasting
• Transformer
• point-wise representation
• patch-wise representation