HINT: Hierarchical Coherent Networks For Constrained Probabilistic Forecasting

Kin G. Olivares, David Luo, Cristian Ignacio Challu, Stefania La Vattiata, Max Mergenthaler Canseco, Artur Dubrawski

Keywords: Time Series, Hierarchical Forecasting, Probabilistic Coherence, Neural Networks, Multivariate Mixture

TL;DR: HINT: Novel distribution/architecture-agnostic framework for hierarchically constrained probabilistic forecasts. Significant 13.2% accuracy improvements over SoTA with strong theoretical foundations.

Abstract:

Large collections of time series data are commonly organized into hierarchies with different levels of aggregation.
We present Hierarchical Coherent Networks (HINT), a forecasting framework that adheres to these aggregation constraints. We specialized HINT in the task via a multivariate mixture optimized with composite likelihood and made coherent via bootstrap reconciliation. Additionally, we robustify the networks to stark time series scale variations, incorporating normalized feature extraction and recomposition of output scales within their architecture. We demonstrate improved accuracy compared to the existing state-of-the-art. We provide ablation studies on our model's components and its solid theoretical foundations. HINT's code is available at this http URL.

Submission Number: 1