This repo implements the experiments for the paper:

Quantile regression with ReLU Networks: Estimators and minimax rates
O.H.M. Padilla, W. Tansey, and Y. Chen
https://arxiv.org/abs/2010.08236

The code in python/neural_model.py implements a deep neural network quantile regression model for both univariate and multivariate responses.

Everything is implemented in python3 with pytorch plus plotting with seaborn --- usual dependencies apply.