TimeMosaic: Temporal Heterogeneity Guided Time Series Forecasting via Adaptive Granularity Patch and Segment-wise Decoding

• URL: http://arxiv.org/abs/2509.19406v3
• Date: Mon, 10 Nov 2025 13:17:12 GMT
• Title: TimeMosaic: Temporal Heterogeneity Guided Time Series Forecasting via Adaptive Granularity Patch and Segment-wise Decoding
• Authors: Kuiye Ding, Fanda Fan, Chunyi Hou, Zheya Wang, Lei Wang, Zhengxin Yang, Jianfeng Zhan,
• Abstract summary: TimeMosaic is a forecasting framework that aims to address temporal heterogeneity.<n>TimeMosaic employs adaptive patch embedding to dynamically adjust granularity according to local information density.<n>Our model trained on the large-scale corpus with 321 billion observations achieves performance competitive with state-of-the-art TSFMs.
• Score: 3.64798801374117
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Multivariate time series forecasting is essential in domains such as finance, transportation, climate, and energy. However, existing patch-based methods typically adopt fixed-length segmentation, overlooking the heterogeneity of local temporal dynamics and the decoding heterogeneity of forecasting. Such designs lose details in information-dense regions, introduce redundancy in stable segments, and fail to capture the distinct complexities of short-term and long-term horizons. We propose TimeMosaic, a forecasting framework that aims to address temporal heterogeneity. TimeMosaic employs adaptive patch embedding to dynamically adjust granularity according to local information density, balancing motif reuse with structural clarity while preserving temporal continuity. In addition, it introduces segment-wise decoding that treats each prediction horizon as a related subtask and adapts to horizon-specific difficulty and information requirements, rather than applying a single uniform decoder. Extensive evaluations on benchmark datasets demonstrate that TimeMosaic delivers consistent improvements over existing methods, and our model trained on the large-scale corpus with 321 billion observations achieves performance competitive with state-of-the-art TSFMs.

Related papers

• MEMTS: Internalizing Domain Knowledge via Parameterized Memory for Retrieval-Free Domain Adaptation of Time Series Foundation Models [51.506429027626005]
  Memory for Time Series (MEMTS) is a lightweight and plug-and-play method for retrieval-free domain adaptation in time series forecasting.<n>Key component of MEMTS is a Knowledge Persistence Module (KPM), which internalizes domain-specific temporal dynamics.<n>This paradigm shift enables MEMTS to achieve accurate domain adaptation with constant-time inference and near-zero latency.
  arXiv  Detail & Related papers  (2026-02-14T14:00:06Z)
• Time2General: Learning Spatiotemporal Invariant Representations for Domain-Generalization Video Semantic Segmentation [9.929390581043334]
  Domain Generalized Video Semantic (DGVSS) is trained on a single labeled driving domain.<n>Time2General achieves a substantial improvement in cross-domain accuracy and temporal stability over prior DGVSS and VSS baselines.
  arXiv  Detail & Related papers  (2026-02-10T10:55:25Z)
• A Decomposition-based State Space Model for Multivariate Time-Series Forecasting [0.0]
  We propose an end-to-end decomposition framework using three parallel deep state space model branches to capture trend, seasonal, and residual components.<n>Across standard benchmarks, DecompSSM outperformed strong baselines, indicating the effectiveness of combining component-wise deep state space models and global context refinement.
  arXiv  Detail & Related papers  (2026-02-05T07:17:08Z)
• UniDiff: A Unified Diffusion Framework for Multimodal Time Series Forecasting [90.47915032778366]
  We propose UniDiff, a unified diffusion framework for multimodal time series forecasting.<n>At its core lies a unified and parallel fusion module, where a single cross-attention mechanism integrates structural information from timestamps and semantic context from texts.<n>Experiments on real-world benchmark datasets across eight domains demonstrate that the proposed UniDiff model achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2025-12-08T05:36:14Z)
• Kairos: Towards Adaptive and Generalizable Time Series Foundation Models [27.076542021368056]
  Time series foundation models (TSFMs) have emerged as a powerful paradigm for time series analysis.<n>We propose Kairos, a flexible TSFM framework that integrates a dynamic patching tokenizer and an instance-adaptive positional embedding.<n>Kairos achieves superior performance with much fewer parameters on two common zero-shot benchmarks.
  arXiv  Detail & Related papers  (2025-09-30T06:02:26Z)
• Multivariate Long-term Time Series Forecasting with Fourier Neural Filter [42.60778405812048]
  We introduce FNF as the backbone and DBD as architecture to provide excellent learning capabilities and optimal learning pathways for spatial-temporal modeling.<n>We show that FNF unifies local time-domain and global frequency-domain information processing within a single backbone that extends naturally to spatial modeling.
  arXiv  Detail & Related papers  (2025-06-10T18:40:20Z)
• Enhancing LLMs for Time Series Forecasting via Structure-Guided Cross-Modal Alignment [12.319685395140862]
  We propose a framework that exploits and aligns the state-transition graph structures shared by time-series and linguistic data as sequential modalities.<n> Experiments on multiple benchmarks demonstrate that SGCMA achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2025-05-19T14:30:41Z)
• Unify and Anchor: A Context-Aware Transformer for Cross-Domain Time Series Forecasting [26.59526791215]
  We identify two key challenges in cross-domain time series forecasting: the complexity of temporal patterns and semantic misalignment.<n>We propose the Unify and Anchor" transfer paradigm, which disentangles frequency components for a unified perspective.<n>We introduce ContexTST, a Transformer-based model that employs a time series coordinator for structured representation.
  arXiv  Detail & Related papers  (2025-03-03T04:11:14Z)
• UmambaTSF: A U-shaped Multi-Scale Long-Term Time Series Forecasting Method Using Mamba [7.594115034632109]
  We propose UmambaTSF, a novel long-term time series forecasting framework. It integrates multi-scale feature extraction capabilities of U-shaped encoder-decoder multilayer perceptrons (MLP) with Mamba's long sequence representation. UmambaTSF achieves state-of-the-art performance and excellent generality on widely used benchmark datasets.
  arXiv  Detail & Related papers  (2024-10-15T04:56:43Z)
• Timer-XL: Long-Context Transformers for Unified Time Series Forecasting [67.83502953961505]
  We present Timer-XL, a causal Transformer for unified time series forecasting.<n>Based on large-scale pre-training, Timer-XL achieves state-of-the-art zero-shot performance.
  arXiv  Detail & Related papers  (2024-10-07T07:27:39Z)
• TimeSiam: A Pre-Training Framework for Siamese Time-Series Modeling [67.02157180089573]
  Time series pre-training has recently garnered wide attention for its potential to reduce labeling expenses and benefit various downstream tasks. This paper proposes TimeSiam as a simple but effective self-supervised pre-training framework for Time series based on Siamese networks.
  arXiv  Detail & Related papers  (2024-02-04T13:10:51Z)
• Rethinking Urban Mobility Prediction: A Super-Multivariate Time Series Forecasting Approach [71.67506068703314]
  Long-term urban mobility predictions play a crucial role in the effective management of urban facilities and services. Traditionally, urban mobility data has been structured as videos, treating longitude and latitude as fundamental pixels. In our research, we introduce a fresh perspective on urban mobility prediction. Instead of oversimplifying urban mobility data as traditional video data, we regard it as a complex time series.
  arXiv  Detail & Related papers  (2023-12-04T07:39:05Z)
• Spatial-Temporal-Decoupled Masked Pre-training for Spatiotemporal Forecasting [15.446085872077898]
  We propose a self-supervised pre-training framework that employs two decoupled masked autoencoders to reconstruct totemporal series along the spatial and temporal dimensions. Rich-context representations learned through such reconstruction could be seamlessly integrated by downstream with predictors arbitrary architectures to augment their performances.
  arXiv  Detail & Related papers  (2023-12-01T11:43:49Z)
• FormerTime: Hierarchical Multi-Scale Representations for Multivariate Time Series Classification [53.55504611255664]
  FormerTime is a hierarchical representation model for improving the classification capacity for the multivariate time series classification task. It exhibits three aspects of merits: (1) learning hierarchical multi-scale representations from time series data, (2) inheriting the strength of both transformers and convolutional networks, and (3) tacking the efficiency challenges incurred by the self-attention mechanism.
  arXiv  Detail & Related papers  (2023-02-20T07:46:14Z)
• Unsupervised Representation Learning for Time Series with Temporal Neighborhood Coding [8.45908939323268]
  We propose a self-supervised framework for learning generalizable representations for non-stationary time series. Our motivation stems from the medical field, where the ability to model the dynamic nature of time series data is especially valuable.
  arXiv  Detail & Related papers  (2021-06-01T19:53:24Z)
• Supporting Optimal Phase Space Reconstructions Using Neural Network Architecture for Time Series Modeling [68.8204255655161]
  We propose an artificial neural network with a mechanism to implicitly learn the phase spaces properties. Our approach is either as competitive as or better than most state-of-the-art strategies.
  arXiv  Detail & Related papers  (2020-06-19T21:04:47Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.