方法对比

	ARIMA（自回归积分滑动平均）模型	FEDformer：频率增强分解Transformer	Informer
领域	计量经济学	深度学习	深度学习
方法族	Regression model	Machine learning	Machine learning
起源年份	2015	2022	2021
提出者	Box & Jenkins (Box-Jenkins methodology)	Tian Zhou et al.	Zhou, H. et al.
类型	Univariate time-series model	Frequency-domain decomposed Transformer for time-series forecasting	Transformer (ProbSparse self-attention)
开创性文献	Box, G. E. P., Jenkins, G. M., Reinsel, G. C. & Ljung, G. M. (2015). Time Series Analysis: Forecasting and Control (5th ed.). Wiley. ISBN: 978-1118675021	Zhou, T., Ma, Z., Wen, Q., Wang, X., Sun, L., & Jin, R. (2022). FEDformer: Frequency enhanced decomposed transformer for long-term series forecasting. ICML.	Zhou, H. et al. (2021). Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting. AAAI.
别名	Box-Jenkins model, ARIMA(p,d,q), ARIMA Modeli	Frequency Enhanced Decomposed Transformer, FED-Transformer, Frequency Domain Transformer, Frekans Tabanlı Ayrıştırılmış Dönüştürücü	Informer — Uzun Dizi Transformer Tahmini, Informer transformer, ProbSparse attention forecaster
相关	5	3	5
摘要	ARIMA is a univariate time-series forecasting model that combines autoregressive, integrated (differencing), and moving-average components to predict a single continuous series from its own past. It is the centrepiece of the Box-Jenkins methodology set out in Box, Jenkins, Reinsel & Ljung's Time Series Analysis (5th ed., 2015).	FEDformer is a Transformer-based architecture for long-term multivariate time-series forecasting, introduced by Zhou et al. at ICML 2022. Its core innovation is the combination of seasonal-trend decomposition with frequency-domain attention: instead of computing full token-to-token attention in the time domain, FEDformer projects queries, keys, and values into the frequency domain via Fourier or wavelet transforms and operates on a randomly selected subset of frequency components, achieving linear complexity while preserving global temporal structure.	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.