Lune

ICLR2025Top-tier venue

TimeMixer++: A General Time Series Pattern Machine for Universal Predictive Analysis

Shiyu Wang, Jiawei Li, Xiaoming Shi, Zhou Ye, Baichuan Mo, Wenze Lin, Shengtong Ju, Zhixuan Chu, Ming Jin

Abstract

Time series analysis plays a critical role in numerous applications, supporting tasks such as forecasting, classification, anomaly detection, and imputation. In this work, we present the time series pattern machine (TSPM), a model designed to excel in a broad range of time series tasks through powerful representation and pattern extraction capabilities. Traditional time series models often struggle to capture universal patterns, limiting their effectiveness across diverse tasks. To address this, we define multiple scales in the time domain and various resolutions in the frequency domain, employing various mixing strategies to extract intricate, task-adaptive time series patterns. Specifically, we introduce TIMEMIXER++, a general-purpose TSPM that processes multi-scale time series using (1) multi-resolution time imaging (MRTI), (2) time image decomposition (TID), (3) multi-scale mixing (MCM), and (4) multi-resolution mixing (MRM) to extract comprehensive temporal patterns. MRTI transforms multi-scale time series into multi-resolution time images, capturing patterns across both temporal and frequency domains. TID leverages dual-axis attention to extract seasonal and trend patterns, while MCM hierarchically aggregates these patterns across scales. MRM adaptively integrates all representations across resolutions. TIMEMIXER++ achieves state-of-the-art performance across 8 time series analytical tasks, consistently surpassing both general-purpose and task-specific models. Our work marks a promising step toward the next generation of TSPMs, paving the way for further advancements in time series analysis.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

54 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting cites Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
  • Efficiently Modeling Long Sequences with Structured State Spaces cites Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
  • xLSTM-Mixer: Multivariate Time Series Forecasting by Mixing via Scalar Memories cites FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting
  • xLSTM-Mixer: Multivariate Time Series Forecasting by Mixing via Scalar Memories cites Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting
  • xLSTM-Mixer: Multivariate Time Series Forecasting by Mixing via Scalar Memories cites Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
  • xLSTM-Mixer: Multivariate Time Series Forecasting by Mixing via Scalar Memories cites Efficiently Modeling Long Sequences with Structured State Spaces
  • xLSTM-Mixer: Multivariate Time Series Forecasting by Mixing via Scalar Memories cites Are Transformers Effective for Time Series Forecasting?
  • OLinear: A Linear Model for Time Series Forecasting in Orthogonally Transformed Domain cites FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting
  • OLinear: A Linear Model for Time Series Forecasting in Orthogonally Transformed Domain cites Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting
  • OLinear: A Linear Model for Time Series Forecasting in Orthogonally Transformed Domain cites Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
  • OLinear: A Linear Model for Time Series Forecasting in Orthogonally Transformed Domain cites Are Transformers Effective for Time Series Forecasting?
  • Are Transformers Effective for Time Series Forecasting? cites FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting
  • Are Transformers Effective for Time Series Forecasting? cites Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting
  • Are Transformers Effective for Time Series Forecasting? cites Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
  • Multi-Scale Finetuning for Encoder-based Time Series Foundation Models cites Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting
  • Multi-Scale Finetuning for Encoder-based Time Series Foundation Models cites Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
  • Multi-Scale Finetuning for Encoder-based Time Series Foundation Models cites Are Transformers Effective for Time Series Forecasting?
  • Aurora: Towards Universal Generative Multimodal Time Series Forecasting cites Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting
  • Aurora: Towards Universal Generative Multimodal Time Series Forecasting cites Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
  • Aurora: Towards Universal Generative Multimodal Time Series Forecasting cites OLinear: A Linear Model for Time Series Forecasting in Orthogonally Transformed Domain
  • Aurora: Towards Universal Generative Multimodal Time Series Forecasting cites Are Transformers Effective for Time Series Forecasting?
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers54

  1. OLinear: A Linear Model for Time Series Forecasting in Orthogonally Transformed DomainWenzhen Yue, Yong Liu, Hao Wang, Haoxuan Li et al.NeurIPS 2025 · 42 citations
  2. Aurora: Towards Universal Generative Multimodal Time Series ForecastingXingjian Wu, Jianxin Jin, Wanghui Qiu, Peng Chen et al.ICLR 2026 · 33 citations
  3. TimeOmni-1: Incentivizing Complex Reasoning with Time Series in Large Language ModelsTong Guan, Zijie Meng, Dianqi Li, Shiyu Wang et al.ICLR 2026 · 29 citations
  4. xLSTM-Mixer: Multivariate Time Series Forecasting by Mixing via Scalar MemoriesMaurice Kraus, Felix Divo, Devendra Singh Dhami, Kristian KerstingNeurIPS 2025 · 27 citations
  5. Multi-Scale Finetuning for Encoder-based Time Series Foundation ModelsZhongzheng Qiao, Chenghao Liu, Yiming Zhang, Ming Jin et al.NeurIPS 2025 · 17 citations

Builds on23

  1. Informer: Beyond Efficient Transformer for Long Sequence Time-Series ForecastingHaoyi Zhou, Shanghang Zhang, Jieqi Peng, Shuai Zhang et al.AAAI 2021 · 7,289 citations
  2. Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series ForecastingHaixu Wu, Jiehui Xu, Jianmin Wang, Mingsheng LongNeurIPS 2021 · 5,824 citations
  3. Are Transformers Effective for Time Series Forecasting?Ailing Zeng, Muxi Chen, Lei Zhang, Qiang XuAAAI 2023 · 3,619 citations
  4. Efficiently Modeling Long Sequences with Structured State SpacesAlbert Gu, Karan Goel, Christopher RéICLR 2022 · 3,482 citations
  5. FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series ForecastingTian Zhou, Ziqing Ma, Qingsong Wen, Xue Wang et al.ICML 2022 · 2,912 citations

Related papers

  1. TimeMixer: Decomposable Multiscale Mixing for Time Series ForecastingShiyu Wang, Haixu Wu, Xiaoming Shi, Tengge Hu et al.ICLR 2024 · 573 citations
  2. GTM: A General Time-series Model for Enhanced Representation Learning of Time-Series dataCheng He, Xu Huang, Gangwei Jiang, Zhaoyi Li et al.ICLR 2026 · 4 citations
  3. Tiny Time Mixers (TTMs): Fast Pre-trained Models for Enhanced Zero/Few-Shot Forecasting of Multivariate Time SeriesVijay Ekambaram, Arindam Jati, Pankaj Dayama, Sumanta Mukherjee et al.NeurIPS 2024 · 207 citations
  4. A Multi-Scale Decomposition MLP-Mixer for Time Series AnalysisShuhan Zhong, Sizhe Song, Weipeng Zhuo, Guanyao Li et al.VLDB 2024 · 48 citations
  5. TimesNet: Temporal 2D-Variation Modeling for General Time Series AnalysisHaixu Wu, Tengge Hu, Yong Liu, Hang Zhou et al.ICLR 2023 · 423 citations
Dusk over the sea between two cliffs drawn in fine vertical lines