# Temporal Fusion Transformers for interpretable multi-horizon time series forecasting

“Temporal Fusion Transformers for interpretable multi-horizon time series forecasting” (2021), by Bryan Y. Lim et al. Computer Science research paper. Original abstract below, unchanged. A free PDF is available.

- Authors: Bryan Y. Lim, Sercan Ö. Arık, Nicolas Loeff, Tomas Pfister
- Published: June 16, 2021
- Field: Computer Science
- DOI: https://doi.org/10.1016/j.ijforecast.2021.03.012
- Free PDF: https://arxiv.org/pdf/1912.09363

## Abstract

Multi-horizon forecasting often contains a complex mix of inputs – including static (i.e. time-invariant) covariates, known future inputs, and other exogenous time series that are only observed in the past – without any prior information on how they interact with the target. Several deep learning methods have been proposed, but they are typically ‘black-box’ models that do not shed light on how they use the full range of inputs present in practical scenarios. In this paper, we introduce the Temporal Fusion Transformer (TFT) – a novel attention-based architecture that combines high-performance multi-horizon forecasting with interpretable insights into temporal dynamics. To learn temporal relationships at different scales, TFT uses recurrent layers for local processing and interpretable self-attention layers for long-term dependencies. TFT utilizes specialized components to select relevant features and a series of gating layers to suppress unnecessary components, enabling high performance in a wide range of scenarios. On a variety of real-world datasets, we demonstrate significant performance improvements over existing benchmarks, and highlight three practical interpretability use cases of TFT.

Canonical page: https://nerdi.pages.dev/paper/W2996331899
Read three real papers a day in Nerdi (iPhone): https://apps.apple.com/app/apple-store/id6813826387?pt=128757794&ct=web-paper&mt=8