Time Series Forecasting

Time Series Forecasting is the task of fitting a model to historical, time-stamped data in order to predict future values. Traditional approaches include moving average, exponential smoothing, and ARIMA, though models as various as RNNs, Transformers, or XGBoost can also be applied. The most popular benchmark is the ETTh1 dataset. Models are typically evaluated using the Mean Square Error (MSE) or Root Mean Square Error (RMSE).

( Image credit: ThaiBinh Nguyen )

Papers

Recently Added Most Hyped Most Active Needs Verification Most Verified

Showing 1–25 of 1609 papers

Title	Date	Tasks	Status	Hype
Logo-LLM: Local and Global Modeling with Large Language Models for Time Series Forecasting	May 16, 2025	Time SeriesTime Series Forecasting	CodeCode Available	7
TimeXer: Empowering Transformers for Time Series Forecasting with Exogenous Variables	Feb 29, 2024	Time SeriesTime Series Forecasting	CodeCode Available	7
A decoder-only foundation model for time-series forecasting	Oct 14, 2023	DecoderTime Series	CodeCode Available	6
iTransformer: Inverted Transformers Are Effective for Time Series Forecasting	Oct 10, 2023	Time SeriesTime Series Forecasting	CodeCode Available	6
SegRNN: Segment Recurrent Neural Network for Long-Term Time Series Forecasting	Aug 22, 2023	Time SeriesTime Series Forecasting	CodeCode Available	6
AutoGluon-TimeSeries: AutoML for Probabilistic Time Series Forecasting	Aug 10, 2023	AutoMLPhilosophy	CodeCode Available	6
BLAST: Balanced Sampling Time Series Corpus for Universal Forecasting Models	May 23, 2025	DiversityTime Series	CodeCode Available	5
Scaling Up Your Kernels: Large Kernel Design in ConvNets towards Universal Representations	Oct 10, 2024	Time Series ForecastingVideo Recognition	CodeCode Available	5
aeon: a Python toolkit for learning from time series	Jun 20, 2024	Anomaly DetectionModel Selection	CodeCode Available	5
TimeMixer: Decomposable Multiscale Mixing for Time Series Forecasting	May 23, 2024	Future predictionTime Series	CodeCode Available	5
TFB: Towards Comprehensive and Fair Benchmarking of Time Series Forecasting Methods	Mar 29, 2024	BenchmarkingMultivariate Time Series Forecasting	CodeCode Available	5
Unified Training of Universal Time Series Forecasting Transformers	Feb 4, 2024	Time SeriesTime Series Forecasting	CodeCode Available	5
UniRepLKNet: A Universal Perception Large-Kernel ConvNet for Audio Video Point Cloud Time-Series and Image Recognition	Jan 1, 2024	Time SeriesTime Series Forecasting	CodeCode Available	5
Long-term Forecasting with TiDE: Time-series Dense Encoder	Apr 17, 2023	Anomaly DetectionDecoder	CodeCode Available	5
A Time Series is Worth 64 Words: Long-term Forecasting with Transformers	Nov 27, 2022	Multivariate Time Series ForecastingRepresentation Learning	CodeCode Available	5
Robust Multivariate Time-Series Forecasting: Adversarial Attacks and Defense Mechanisms	Jul 19, 2022	Adversarial AttackMultivariate Time Series Forecasting	CodeCode Available	5
Timer-XL: Long-Context Transformers for Unified Time Series Forecasting	Oct 7, 2024	Time SeriesTime Series Forecasting	CodeCode Available	4
Zero-shot forecasting of chaotic systems	Sep 24, 2024	AttributeIn-Context Learning	CodeCode Available	4
Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts	Sep 24, 2024	Computational EfficiencyMixture-of-Experts	CodeCode Available	4
UniTS: A Unified Multi-Task Time Series Model	Feb 29, 2024	Anomaly DetectionImputation	CodeCode Available	4
Timer: Generative Pre-trained Transformers Are Large Time Series Models	Feb 4, 2024	Anomaly DetectionImputation	CodeCode Available	4
Tiny Time Mixers (TTMs): Fast Pre-trained Models for Enhanced Zero/Few-Shot Forecasting of Multivariate Time Series	Jan 8, 2024	CPUFew-Shot Learning	CodeCode Available	4
Time-LLM: Time Series Forecasting by Reprogramming Large Language Models	Oct 3, 2023	Time SeriesTime Series Forecasting	CodeCode Available	4
Are Transformers Effective for Time Series Forecasting?	May 26, 2022	Anomaly DetectionRelation Extraction	CodeCode Available	4
Efficient Automated Deep Learning for Time Series Forecasting	May 11, 2022	AutoMLBayesian Optimization	CodeCode Available	4

Show:10 25 50

← PrevPage 1 of 65Next →

All datasets ETTh1 (336) Multivariate ETTh1 (720) Multivariate ETTh2 (336) Multivariate ETTh2 (720) Multivariate ETTh1 (192) Multivariate ETTh2 (192) Multivariate ETTh2 (96) Multivariate ETTh1 (96) Multivariate Weather (192)ETTh1 (720) Univariate Weather (96)Electricity (96)

Benchmark Results

#	Model	Metric	Claimed	Verified	Status
1	Informer	MSE	0.88	—	Unverified
2	QuerySelector	MSE	0.85	—	Unverified
3	Transformer	MSE	0.83	—	Unverified
4	Aaren	MSE	0.65	—	Unverified
5	RPMixer	MSE	0.52	—	Unverified
6	ATFNet	MSE	0.51	—	Unverified
7	MOIRAILarge	MSE	0.51	—	Unverified
8	Autoformer	MSE	0.51	—	Unverified
9	SCINet	MSE	0.5	—	Unverified
10	S-Mamba	MSE	0.49	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	QuerySelector	MSE	1.12	—	Unverified
2	Transformer	MSE	1.11	—	Unverified
3	Informer	MSE	0.94	—	Unverified
4	GLinear	MSE	0.59	—	Unverified
5	SCINet	MSE	0.54	—	Unverified
6	MoLE-DLinear	MSE	0.51	—	Unverified
7	PRformer	MSE	0.49	—	Unverified
8	TEFN	MSE	0.48	—	Unverified
9	DLinear	MSE	0.47	—	Unverified
10	FiLM	MSE	0.47	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	Transformer	MSE	2.66	—	Unverified
2	QuerySelector	MSE	2.32	—	Unverified
3	Informer	MSE	1.67	—	Unverified
4	DLinear	MSE	0.45	—	Unverified
5	TEFN	MSE	0.42	—	Unverified
6	MoLE-DLinear	MSE	0.42	—	Unverified
7	FiLM	MSE	0.38	—	Unverified
8	MoLE-RLinear	MSE	0.37	—	Unverified
9	SCINet	MSE	0.37	—	Unverified
10	PRformer	MSE	0.36	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	Transformer	MSE	3.18	—	Unverified
2	QuerySelector	MSE	3.07	—	Unverified
3	Informer	MSE	2.34	—	Unverified
4	MoLE-DLinear	MSE	0.61	—	Unverified
5	DLinear	MSE	0.61	—	Unverified
6	SCINet	MSE	0.48	—	Unverified
7	FiLM	MSE	0.44	—	Unverified
8	TEFN	MSE	0.43	—	Unverified
9	TiDE	MSE	0.42	—	Unverified
10	MoLE-RLinear	MSE	0.41	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	MoLE-DLinear	MSE	0.45	—	Unverified
2	TEFN	MSE	0.43	—	Unverified
3	FiLM	MSE	0.41	—	Unverified
4	PatchTST/64	MSE	0.41	—	Unverified
5	TiDE	MSE	0.41	—	Unverified
6	NLinear	MSE	0.41	—	Unverified
7	DiPE-Linear	MSE	0.41	—	Unverified
8	DLinear	MSE	0.41	—	Unverified
9	RLinear	MSE	0.4	—	Unverified
10	MoLE-RLinear	MSE	0.4	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	DLinear	MSE	0.38	—	Unverified
2	TEFN	MSE	0.38	—	Unverified
3	MoLE-DLinear	MSE	0.36	—	Unverified
4	FiLM	MSE	0.36	—	Unverified
5	NLinear	MSE	0.34	—	Unverified
6	PatchTST/64	MSE	0.34	—	Unverified
7	MoLE-RLinear	MSE	0.34	—	Unverified
8	TiDE	MSE	0.33	—	Unverified
9	PRformer	MSE	0.33	—	Unverified
10	LTBoost (drop_last=false)	MSE	0.33	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	DLinear	MSE	0.29	—	Unverified
2	TEFN	MSE	0.29	—	Unverified
3	MoLE-DLinear	MSE	0.29	—	Unverified
4	FiLM	MSE	0.28	—	Unverified
5	NLinear	MSE	0.28	—	Unverified
6	TSMixer	MSE	0.28	—	Unverified
7	DiPE-Linear	MSE	0.28	—	Unverified
8	PatchTST/64	MSE	0.27	—	Unverified
9	MoLE-RLinear	MSE	0.27	—	Unverified
10	TiDE	MSE	0.27	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	TEFN	MSE	0.38	—	Unverified
2	MoLE-DLinear	MSE	0.38	—	Unverified
3	TiDE	MSE	0.38	—	Unverified
4	MoLE-RLinear	MSE	0.38	—	Unverified
5	FiLM	MSE	0.37	—	Unverified
6	PatchTST/64	MSE	0.37	—	Unverified
7	DiPE-Linear	MSE	0.37	—	Unverified
8	TSMixer	MSE	0.37	—	Unverified
9	RLinear	MSE	0.37	—	Unverified
10	TTM	MSE	0.36	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	TEFN	MSE	0.23	—	Unverified
2	DLinear	MSE	0.22	—	Unverified