Optimal number of clusters in explainable data analysis of agent-based simulation experiments

Open Webpage

Shengkun Xie, Anna T. Lawniczak, Chong Gan

Published: 2022, Last Modified: 30 Sept 2025J. Comput. Sci. 2022EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Two novel approaches are proposed to find the optimal number of clusters in K-mean clustering.•The methods are designed to meet the need of analysis of agent-based computer simulations exclusively for the purpose of making the results more interpretable.•The proposed methods help improve understanding of the causality of input parameters and the response variables of the complex systems that we study.•Analysis related to agent-based simulation is addressed in the domain of data clustering.