Saved in:
Bibliographic Details
Main Authors: Lang, Quanjun, Wang, Xiong, Lu, Fei, Maggioni, Mauro
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.03954
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917246927896576
author Lang, Quanjun
Wang, Xiong
Lu, Fei
Maggioni, Mauro
author_facet Lang, Quanjun
Wang, Xiong
Lu, Fei
Maggioni, Mauro
contents We propose a framework for the joint inference of network topology, multi-type interaction kernels, and latent type assignments in heterogeneous interacting particle systems from multi-trajectory data. This learning task is a challenging non-convex mixed-integer optimization problem, which we address through a novel three-stage approach. First, we leverage shared structure across agent interactions to recover a low-rank embedding of the system parameters via matrix sensing. Second, we identify discrete interaction types by clustering within the learned embedding. Third, we recover the network weight matrix and kernel coefficients through matrix factorization and a post-processing refinement. We provide theoretical guarantees with estimation error bounds under a Restricted Isometry Property (RIP) assumption and establish conditions for the exact recovery of interaction types based on cluster separability. Numerical experiments on synthetic datasets, including heterogeneous predator-prey systems, demonstrate that our method yields an accurate reconstruction of the underlying dynamics and is robust to noise.
format Preprint
id arxiv_https___arxiv_org_abs_2602_03954
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Learning Multi-type heterogeneous interacting particle systems
Lang, Quanjun
Wang, Xiong
Lu, Fei
Maggioni, Mauro
Machine Learning
Computation
Methodology
We propose a framework for the joint inference of network topology, multi-type interaction kernels, and latent type assignments in heterogeneous interacting particle systems from multi-trajectory data. This learning task is a challenging non-convex mixed-integer optimization problem, which we address through a novel three-stage approach. First, we leverage shared structure across agent interactions to recover a low-rank embedding of the system parameters via matrix sensing. Second, we identify discrete interaction types by clustering within the learned embedding. Third, we recover the network weight matrix and kernel coefficients through matrix factorization and a post-processing refinement. We provide theoretical guarantees with estimation error bounds under a Restricted Isometry Property (RIP) assumption and establish conditions for the exact recovery of interaction types based on cluster separability. Numerical experiments on synthetic datasets, including heterogeneous predator-prey systems, demonstrate that our method yields an accurate reconstruction of the underlying dynamics and is robust to noise.
title Learning Multi-type heterogeneous interacting particle systems
topic Machine Learning
Computation
Methodology
url https://arxiv.org/abs/2602.03954