Algorithms support for time-series forecasting - Amazon SageMaker

Algorithms support for time-series forecasting

Autopilot trains the following six built-in algorithms with your target time-series. Then, using a stacking ensemble method, it combines these model candidates to create an optimal forecasting model for a given objective metric.

  • Convolutional Neural Network - Quantile Regression (CNN-QR) – CNN-QR is a proprietary machine learning algorithm for forecasting time-series using causal convolutional neural networks (CNNs). CNN-QR works best with large datasets containing hundreds of time-series.

  • DeepAR+ – DeepAR+ is a proprietary machine learning algorithm for forecasting time-series using recurrent neural networks (RNNs). DeepAR+ works best with large datasets containing hundreds of feature time-series.

  • Prophet – Prophet is a popular local Bayesian structural time series model based on an additive model where non-linear trends are fit with yearly, weekly, and daily seasonality. The Autopilot Prophet algorithm uses the Prophet class of the Python implementation of Prophet. It works best with time-series with strong seasonal effects and several seasons of historical data.

  • Non-Parametric Time Series (NPTS) – The NPTS proprietary algorithm is a scalable, probabilistic baseline forecaster. It predicts the future value distribution of a given time-series by sampling from past observations. NPTS is especially useful when working with sparse or intermittent time series.

  • Autoregressive Integrated Moving Average (ARIMA) – ARIMA is a commonly used statistical algorithm for time-series forecasting. The algorithm captures standard temporal structures (patterned organizations of time) in the input dataset. It is especially useful for simple datasets with under 100 time series.

  • Exponential Smoothing (ETS) – ETS is a commonly used statistical algorithm for time-series forecasting. The algorithm is especially useful for simple datasets with under 100 time series, and datasets with seasonality patterns. ETS computes a weighted average over all observations in the time series dataset as its prediction, with exponentially decreasing weights over time.