Saved in:
Bibliographic Details
Main Authors: Zeng, Zhijun, Chen, Junqing, Shi, Zuoqiang
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.12316
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908656466919424
author Zeng, Zhijun
Chen, Junqing
Shi, Zuoqiang
author_facet Zeng, Zhijun
Chen, Junqing
Shi, Zuoqiang
contents We study an inverse problem for stochastic and quantum dynamical systems in a time-label-free setting, where only unordered density snapshots sampled at unknown times drawn from an observation-time distribution are available. These observations induce a distribution over state densities, from which we seek to recover the parameters of the underlying evolution operator. We formulate this as learning a distribution-to-function neural operator and propose BlinDNO, a permutation-invariant architecture that integrates a multiscale U-Net encoder with an attention-based mixer. Numerical experiments on a wide range of stochastic and quantum systems, including a 3D protein-folding mechanism reconstruction problem in a cryo-EM setting, demonstrate that BlinDNO reliably recovers governing parameters and consistently outperforms existing neural inverse operator baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2511_12316
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle BlinDNO: A Distributional Neural Operator for Dynamical System Reconstruction from Time-Label-Free data
Zeng, Zhijun
Chen, Junqing
Shi, Zuoqiang
Machine Learning
Computational Engineering, Finance, and Science
Dynamical Systems
35R30, 65M32, 68T07
We study an inverse problem for stochastic and quantum dynamical systems in a time-label-free setting, where only unordered density snapshots sampled at unknown times drawn from an observation-time distribution are available. These observations induce a distribution over state densities, from which we seek to recover the parameters of the underlying evolution operator. We formulate this as learning a distribution-to-function neural operator and propose BlinDNO, a permutation-invariant architecture that integrates a multiscale U-Net encoder with an attention-based mixer. Numerical experiments on a wide range of stochastic and quantum systems, including a 3D protein-folding mechanism reconstruction problem in a cryo-EM setting, demonstrate that BlinDNO reliably recovers governing parameters and consistently outperforms existing neural inverse operator baselines.
title BlinDNO: A Distributional Neural Operator for Dynamical System Reconstruction from Time-Label-Free data
topic Machine Learning
Computational Engineering, Finance, and Science
Dynamical Systems
35R30, 65M32, 68T07
url https://arxiv.org/abs/2511.12316