方法对比

	非平稳Transformer	增广迪基-福勒（ADF）单位根检验	Informer
领域	深度学习	计量经济学	深度学习
方法族	Machine learning	Regression model	Machine learning
起源年份	2022	1979	2021
提出者	Yong Liu et al.	David A. Dickey & Wayne A. Fuller	Zhou, H. et al.
类型	Transformer-based time-series forecasting model	Unit-root test for stationarity	Transformer (ProbSparse self-attention)
开创性文献	Liu, Y., Wu, H., Wang, J., & Long, M. (2022). Non-stationary transformers: Exploring the stationarity in time series forecasting. NeurIPS.	Dickey, D. A., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association, 74(366a), 427–431.	Zhou, H. et al. (2021). Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting. AAAI.
别名	NS-Transformer, Non-stationary Transformer Network, Stationarization-based Transformer, Durağan-Olmayan Transformer	ADF test, Dickey-Fuller test, unit root test, Genişletilmiş Dickey-Fuller testi	Informer — Uzun Dizi Transformer Tahmini, Informer transformer, ProbSparse attention forecaster
相关	3	4	5
摘要	Non-stationary Transformer is a Transformer-based time-series forecasting architecture introduced by Yong Liu, Haixu Wu, Jianmin Wang, and Mingsheng Long at NeurIPS 2022. It addresses a fundamental tension in applying Transformers to real-world time series: over-stationarization during preprocessing strips out non-stationary signals that carry predictive information, while raw non-stationary inputs cause attention to collapse. The model resolves this through series stationarization paired with a novel de-stationary attention mechanism that restores the original temporal distribution in predictions.	The Augmented Dickey-Fuller (ADF) test is the most widely used test for a unit root — that is, for whether a time series is non-stationary and must be differenced before modelling. Introduced by David Dickey and Wayne Fuller in 1979 and extended by Said and Dickey in 1984 to series with higher-order autocorrelation, it regresses the change in the series on its lagged level plus lagged differences and asks whether the lagged-level coefficient is zero.	Informer is a Transformer-based model introduced by Zhou et al. in 2021 for long-sequence time-series forecasting, using a ProbSparse self-attention mechanism that lowers the computational complexity of the standard Transformer to O(L log L). It is built for problems that demand predictions across thousands of future steps.