DeepAR vs DeepVAR confusion

DeepAR and DeepVAR are booth deep learning models designed for time series forecasting, and they can handle multiple related variables simultaneously. This difference is in their underlying architecture and modelling approach.

The Amazon SageMaker DeepAR forecasting algorithm is a supervised learning algorithm for forecasting scalar (one-dimensional) time series using recurrent neural networks (RNN). Classical forecasting methods, such as autoregressive integrated moving average (ARIMA) or exponential smoothing (ETS), fit a single model to each individual time series. They then use that model to extrapolate the time series into the future. In many applications, however, you have many similar time series across a set of cross-sectional units. For example, you might have time series groupings for demand for different products, server loads, and requests for webpages. For this type of application, you can benefit from training a single model jointly over all of the time series. DeepAR takes this approach. When your dataset contains hundreds of related time series, DeepAR outperforms the standard ARIMA and ETS methods. You can also use the trained model to generate forecasts for new time series that are similar to the ones it has been trained on. https://docs.aws.amazon.com/sagemaker/latest/dg/deepar.html

DeepVAR: DeepVAR, on the other hand, is specifically designed for multivariate time series forecasting. It considers the interdependencies and cross-correlations among multiple variables. DeepVAR is suitable when the relationships between variables are important and need to be explicitly modeled. This model is particularly useful in domains where multiple related variables influence each other, such as macroeconomic forecasting, supply chain management, and portfolio risk analysis. DeepAR is based on an auto-regressive recurrent neural network model, specifically designed for multivariate time-series modeling producing accurate probabilistic forecasts. DeepVAR captures both the temporal dynamics within each variable and the cross-series interactions. DeepVAR is based on the concept of vector autoregression a statical model commonly used in econometrics.