Abstract
Membrane computing also called P system, seeks to discover new computational models from the study of cellular membranes. In this study, we reported our initial efforts to classify Macao visitor expenditure profile using a membrane computing approach. Specifically, we designed a novel P system including specific membrane structure and membrane rules to realize an improved k-medoids clustering algorithm. We also empirically evaluated the new P system with a dataset from a large-scale visitor profile survey supported by Macao Government Tourist Office. Results reveal that the new P system is more robust to initial centers and achieves higher clustering accuracy as compared to the classical k-medoids clustering and the classical k-means method. In addition, we analyzed the demographic and behavior profile of the extracted six expenditure clusters and discussed the theoretical and practical implications of this study.
Original language | English |
---|---|
Title of host publication | 2017 International Conference on Service Systems and Service Management, ICSSSM |
Number of pages | 6 |
ISBN (Electronic) | 978-1-5090-6370-3 |
DOIs | |
Publication status | Published (in print/issue) - Jun 2017 |