N-BEATS: Neural Basis Expansion Analysis for Time Series
Deep neural architecture specifically designed for univariate time series forecasting with backward and forward residual links.
Verdict
Pure deep learning approach with no feature engineering. Impressive results on M4 competition. Good baseline for financial forecasting.
Practical Use
Ideal for univariate financial time series like stock prices, FX rates, or economic indicators where you want strong forecasting performance without manual feature engineering. The interpretable variant can decompose predictions into trend and seasonality components. Note: ServiceNow's official implementation is non-commercial (CC BY-NC 4.0); for commercial use, consider the MIT-licensed nbeats-pytorch package or implement from the paper.
Paper
https://arxiv.org/abs/1905.10437Boris N. Oreshkin, Dmitri Carpov, Nicolas Chapados, Yoshua Bengio. ICLR, 2020
Code Repository
https://github.com/ServiceNow/N-BEATSLicense
Custom. Code (ServiceNow/N-BEATS): CC BY-NC 4.0, non-commercial only; commercial use needs a separate license. Paper: arXiv non-exclusive distribution license (free to read; no reuse license).
Installation
pip install nbeats-pytorch