Dynamic Matching Utilizing Latent Factor Modeling

Zifei Guo, Xiaowu Dai

Keywords: Dynamic matching, Learning, Two-sided market, Bi-clustering, Label aggregation

TL;DR: A latent factor model is introduced to study learning and dynamic matching in two-sided market.

Abstract: This paper investigates the supply-demand matching problem on dynamic platforms, focusing on optimizing matching strategies by learning workers' attributes when their types are uncertain and constantly changing. To address this problem, we introduce a latent factor model and a multi-centroid grouping penalty mechanism to predict latent factors of workers and perform dynamic matching. Our approach operates in two stages: the first stage fits latent feature vectors for workers and jobs and groups them using historical data; the second stage utilizes these latent features for dynamic matching. Our research demonstrates that the introduced model can adapt to the dynamic changes of the platform with good predictive consistency and group robustness, and improves overall operational benefit through continuous optimization of matching results. We provide simulation experiments and a real case study using kidney exchange data and compare our model with a point process model to show that our approach performs well on dynamic platform matching problems.

Primary Area: other topics in machine learning (i.e., none of the above)

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 10597