השוואת שיטות

	DeepAR	מודל ARIMA (Autoregressive Integrated Moving Average)	PatchTST
תחום	למידה עמוקה	אקונומטריקה	למידה עמוקה
משפחה	Machine learning	Regression model	Machine learning
שנת המקור	2020	2015	2023
הוגה השיטה	Salinas, D., Flunkert, V. & Gasthaus, J. (Amazon)	Box & Jenkins (Box-Jenkins methodology)	Nie, Y. et al.
סוג	Autoregressive recurrent neural network (probabilistic forecasting)	Univariate time-series model	Transformer for time series forecasting
מקור מכונן	Salinas, D., Flunkert, V., Gasthaus, J. & Januschowski, T. (2020). DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks. International Journal of Forecasting, 36(3), 1181–1191.	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	Nie, Y., Nguyen, N. H., Sinthong, P. & Kalagnanam, J. (2023). A Time Series is Worth 64 Words: Long-term Forecasting with Transformers. ICLR.
כינויים	DeepAR — Olasılıksal RNN Tahmini, probabilistic autoregressive RNN forecasting, Amazon DeepAR	Box-Jenkins model, ARIMA(p,d,q), ARIMA Modeli	PatchTST — Yama Tabanlı Zaman Serisi Transformer, patch-based time series transformer, channel-independent transformer
קשורות	5	5	3
תקציר	DeepAR is Amazon's industrial forecasting model, introduced by Salinas, Flunkert and Gasthaus (2017; published 2020), that uses an autoregressive recurrent neural network to estimate the parameters of a probability distribution at each step, producing a confidence interval rather than a single point forecast. It can model many related time series jointly within one model.	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).	PatchTST is a patch-based Transformer architecture for time series forecasting, introduced by Nie and colleagues in 2023, that cuts each series into overlapping patches treated as tokens and processes channels independently. It balances computational efficiency with strong accuracy on long-horizon forecasting.