Pith. sign in

REVIEW 19 cited by

Are Transformers Effective for Time Series Forecasting?

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2205.13504 v3 pith:KKKHDYCK submitted 2022-05-26 cs.AI cs.LG

classification cs.AIcs.LG
keywords ltsfseriestimeltsf-linearmodelstemporaltransformer-basedtransformers
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Recently, there has been a surge of Transformer-based solutions for the long-term time series forecasting (LTSF) task. Despite the growing performance over the past few years, we question the validity of this line of research in this work. Specifically, Transformers is arguably the most successful solution to extract the semantic correlations among the elements in a long sequence. However, in time series modeling, we are to extract the temporal relations in an ordered set of continuous points. While employing positional encoding and using tokens to embed sub-series in Transformers facilitate preserving some ordering information, the nature of the \emph{permutation-invariant} self-attention mechanism inevitably results in temporal information loss. To validate our claim, we introduce a set of embarrassingly simple one-layer linear models named LTSF-Linear for comparison. Experimental results on nine real-life datasets show that LTSF-Linear surprisingly outperforms existing sophisticated Transformer-based LTSF models in all cases, and often by a large margin. Moreover, we conduct comprehensive empirical studies to explore the impacts of various design elements of LTSF models on their temporal relation extraction capability. We hope this surprising finding opens up new research directions for the LTSF task. We also advocate revisiting the validity of Transformer-based solutions for other time series analysis tasks (e.g., anomaly detection) in the future. Code is available at: \url{https://github.com/cure-lab/LTSF-Linear}.

Discussion (0). Sign in to comment.

Forward citations

Cited by 19 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. OpenAlex reports about 184 citations worldwide. Full citation record

  1. Everything You Need to Know About CS Education: Open Results from a Survey of More Than 18,000 Participants

    cs.CY 2025-08 conditional novelty 7.0 of 10

    FlowState, an 18.6M-parameter SSM with a functional-basis decoder, reaches top GIFT-Eval MASE and CRPS and generalizes to sampling rates not seen in training.

  2. Agoran: An Agentic Open Marketplace for 6G RAN Automation

    cs.NI 2025-08 unverdicted novelty 7.0 of 10

    Agoran uses AI agents with legislative, executive, and judicial branches to automatically negotiate and manage 6G network slices, achieving large performance gains on a 5G testbed.

  3. Co4ICF: Co-evolving Physics-Informed Surrogate and RL-based Pulse Optimizer for Inertial Confinement Fusion

    cs.AI 2026-07 conditional novelty 6.0 of 10

    A co-evolving physics-informed surrogate and PPO optimizer reaches 146% (1D) and 247% (2D) normalized ICF yield versus baseline without 2D training data.

  4. Flow-PIN: A Two-Stage Power-Flow-Guided Method for System-Wide Multivariate Profile Inpainting in Distribution Networks

    eess.SP 2026-07 conditional novelty 6.0 of 10

    A two-stage framework combining conditional flow matching with a topology-aware power-flow-guided refiner improves the accuracy and physical consistency of missing power grid data imputation.

  5. HAMON: Passive Optical Sequence Mixing for Long-Horizon Forecasting

    cs.LG 2026-06 conditional novelty 6.0 of 10

    A simulated passive diffractive optical core, trained with digital backpropagation, matches or beats strong digital forecasters on several long-horizon benchmarks.

  6. Rethinking Multimodal Fusion for Time Series: Text Modalities Need Constrained Fusion

    cs.LG 2026-03 unverdicted novelty 6.0 of 10

    Uncontrolled text–time-series fusion underperforms unimodal baselines; constrained fusion and a low-rank Controlled Fusion Adapter recover gains without changing the TS backbone.

  7. Time Series Foundation Models for Multivariate Financial Time Series Forecasting

    q-fin.GN 2025-07 reject novelty 6.0 of 10

    Pretrained TTM shows large transfer and sample-efficiency gains in three financial forecasting tasks relative to training from scratch, but methodological flaws including possible look-ahead bias weaken the quantitati...

  8. Decomposing the Time Series Forecasting Pipeline: A Modular Approach for Time Series Representation, Information Extraction, and Projection

    cs.AI 2025-07 conditional novelty 6.0 of 10

    REP-Net, a modular pipeline of representation, memory, and projection modules, achieves competitive forecasting accuracy on seven multivariate benchmarks with lower computational cost.

  9. Neural Functions for Learning Periodic Signal

    cs.LG 2025-06 conditional novelty 6.0 of 10

    NeRT factorizes periodic signals into a sine-based periodic factor and an unbounded scale factor, enabling extrapolation beyond the training range on several periodic benchmarks.

  10. NSW-EPNews: A News-Augmented Benchmark for Electricity Price Forecasting with LLMs

    cs.LG 2025-05 reject novelty 6.0 of 10

    LLMs forecast electricity prices worse than ARIMA on the new NSW-EPNews benchmark and frequently hallucinate by echoing, offsetting, or repeating historical prices.

  11. MoTime: A Dataset Suite for Multimodal Time Series Forecasting

    cs.LG 2025-05 conditional novelty 6.0 of 10

    MoTime provides a large multimodal forecasting benchmark and shows that external text or images can improve forecasts in some datasets, especially cold-start and sparse settings, though gains are inconsistent.

  12. RhyMix: A Lightweight Adaptive Multi-Rhythm Network for Long-Term Time Series Forecasting

    cs.LG 2026-07 conditional novelty 5.5 of 10

    RhyMix reaches state-of-the-art long-term multivariate forecasting on 10 of 12 public benchmarks with a ~40K-parameter dual-path adaptive architecture of linear complexity.

  13. MSTN: A Lightweight and Fast Model for General TimeSeries Analysis

    cs.LG 2025-11 unverdicted novelty 5.0 of 10

    MSTN is a lightweight hybrid model that reports new state-of-the-art results on 33 of 40 time series benchmarks for imputation, forecasting, and classification while using under one million parameters and sub-second i...

  14. BALM-TSF: Balanced Multimodal Alignment for LLM-Based Time Series Forecasting

    cs.AI 2025-08 conditional novelty 5.0 of 10

    BALM-TSF combines a statistical-prompt text branch with a patch-based time series branch, using scaling plus contrastive alignment to balance the two modalities, improving long-term and few-shot forecasting on five of...

  15. Benchmarking Unsupervised Strategies for Anomaly Detection in Multivariate Time Series

    cs.LG 2025-06 conditional novelty 5.0 of 10

    Across ten public datasets, a reconstruction-based inverted transformer with per-variate anomaly labelling achieves the best or tied best MCC on most datasets, but the comparison is weakened by test-set-based configur...

  16. Challenges of Explainability in Continual Learning for Time Series Forecasting

    cs.LG 2026-07 conditional novelty 4.0 of 10

    On two French groundwater time series, attention-based replay sampling looks more structured and interpretable than random or loss-based alternatives, but the paper reports no numeric metrics to support its performanc...

  17. Foundation vs. Specialized Models: Evaluating Catastrophic Forgetting in Continual Time Series Forecasting

    cs.LG 2025-10 reject novelty 4.0 of 10

    Fine-tuning TimesFM sequentially on new synthetic time-series data causes measurable forgetting of earlier tasks, with higher learning rates producing stronger forgetting.

  18. Synthetic Time Series Forecasting with Transformer Architectures: Extensive Simulation Benchmarks

    cs.LG 2025-05 reject novelty 4.0 of 10

    Autoformer and PatchTST outperform Informer across synthetic forecasting benchmarks, while a proposed Koopman-Transformer hybrid is illustrated on Van der Pol and Lorenz systems.

  19. Foundation Models for Clean Energy Forecasting: A Comprehensive Review

    eess.SY 2025-07 conditional novelty 3.0 of 10

    A survey of foundation model methods, data, and open problems for renewable energy forecasting, built from roughly 218 cited works.

Pith tools