Preference-Driven Spatial-Temporal Counting Process Models
Keywords: choice model, spatial-temporal counting process model
Abstract: Traditional spatial-temporal models often overlook the complex decision-making processes and social factors that shape spatial-temporal event data generated by humans. This paper introduces a novel framework that integrates choice theory with social intelligence to model and analyze counting processes, such as crime occurrences or bike-sharing activity, where the observed discrete events result from individual decisions influenced by social dynamics.
Our approach aims to uncover latent human preference patterns, represented by utility functions, to capture the diverse decision-making factors within a population that result in the observed event counts. These latent factors help explain how choices—such as where and when to commit a crime—are shaped by personal preferences, environmental conditions, and social influences. By modeling the aggregate outcomes of these individual choices, we can better understand and predict patterns in counting processes. The proposed model adopts a preference-driven approach to counting data, providing interpretable insights at a detailed level. It also enables in-depth analysis of how external interventions, like law enforcement actions or policy changes, influence individual decisions and how these effects spread through the system. Empirical evaluation of crime and bike-sharing datasets demonstrates our model's ability to offer clear insights and achieve high predictive accuracy.
Primary Area: interpretability and explainable AI
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: 13442
Loading