Deep Distributional Learning with Non-crossing Quantile Network

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shen, Guohao, Dai, Runpeng, Wu, Guojun, Luo, Shikai, Shi, Chengchun, Zhu, Hongtu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910908995862528
author Shen, Guohao
Dai, Runpeng
Wu, Guojun
Luo, Shikai
Shi, Chengchun
Zhu, Hongtu
author_facet Shen, Guohao
Dai, Runpeng
Wu, Guojun
Luo, Shikai
Shi, Chengchun
Zhu, Hongtu
contents In this paper, we introduce a non-crossing quantile (NQ) network for conditional distribution learning. By leveraging non-negative activation functions, the NQ network ensures that the learned distributions remain monotonic, effectively addressing the issue of quantile crossing. Furthermore, the NQ network-based deep distributional learning framework is highly adaptable, applicable to a wide range of applications, from classical non-parametric quantile regression to more advanced tasks such as causal effect estimation and distributional reinforcement learning (RL). We also develop a comprehensive theoretical foundation for the deep NQ estimator and its application to distributional RL, providing an in-depth analysis that demonstrates its effectiveness across these domains. Our experimental results further highlight the robustness and versatility of the NQ network.
format Preprint
id arxiv_https___arxiv_org_abs_2504_08215
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Deep Distributional Learning with Non-crossing Quantile Network
Shen, Guohao
Dai, Runpeng
Wu, Guojun
Luo, Shikai
Shi, Chengchun
Zhu, Hongtu
Machine Learning
Statistics Theory
In this paper, we introduce a non-crossing quantile (NQ) network for conditional distribution learning. By leveraging non-negative activation functions, the NQ network ensures that the learned distributions remain monotonic, effectively addressing the issue of quantile crossing. Furthermore, the NQ network-based deep distributional learning framework is highly adaptable, applicable to a wide range of applications, from classical non-parametric quantile regression to more advanced tasks such as causal effect estimation and distributional reinforcement learning (RL). We also develop a comprehensive theoretical foundation for the deep NQ estimator and its application to distributional RL, providing an in-depth analysis that demonstrates its effectiveness across these domains. Our experimental results further highlight the robustness and versatility of the NQ network.
title Deep Distributional Learning with Non-crossing Quantile Network
topic Machine Learning
Statistics Theory
url https://arxiv.org/abs/2504.08215