Deep distribution regression

Li, Rui; Reich, Brian J.; Bondell, Howard D.

Deep distribution regression

Embargo End Date

2023-02-01

Type

Article

Authors

Li, Rui

Reich, Brian J.
Bondell, Howard D.

KAUST Grant Number

3800.2

Online Publication Date

2021-02-22

Print Publication Date

2021-07

Date

2021-02-22

Abstract

Due to their flexibility and predictive performance, machine-learning based regression methods have become an important tool for predictive modeling and forecasting. However, most methods focus on estimating the conditional mean or specific quantiles of the target quantity and do not provide the full conditional distribution, which contains uncertainty information that might be crucial for decision making. A general solution consists of transforming a conditional distribution estimation problem into a constrained multi-class classification problem, in which tools such as deep neural networks can be applied. A novel joint binary cross-entropy loss function is proposed to accomplish this goal. Its performance is compared to current state-of-the-art methods via simulation. The approach also shows improved accuracy in a probabilistic solar energy forecasting problem.

Citation

Li, R., Reich, B. J., & Bondell, H. D. (2021). Deep distribution regression. Computational Statistics & Data Analysis, 107203. doi:10.1016/j.csda.2021.107203

Acknowledgements

The authors’ work was partially supported by King Abdullah University of Science and Technology (grant number 3800.2).