Interpretable Feature Engineering for Time Series Predictors using Attention Networks
This work addresses interpretable feature engineering for time-series predictors, particularly in banking and other applications, but is incremental as it builds on existing attention mechanisms.
The paper tackles regression problems with time-series predictors by using multi-head attention networks to develop interpretable features, achieving good predictive performance as demonstrated through simulation and real datasets.
Regression problems with time-series predictors are common in banking and many other areas of application. In this paper, we use multi-head attention networks to develop interpretable features and use them to achieve good predictive performance. The customized attention layer explicitly uses multiplicative interactions and builds feature-engineering heads that capture temporal dynamics in a parsimonious manner. Convolutional layers are used to combine multivariate time series. We also discuss methods for handling static covariates in the modeling process. Visualization and explanation tools are used to interpret the results and explain the relationship between the inputs and the extracted features. Both simulation and real dataset are used to illustrate the usefulness of the methodology. Keyword: Attention heads, Deep neural networks, Interpretable feature engineering