Identifying patients in target customer segments using a two-stage clustering-classification approach: A hospital-based assessment

  • Authors:

  • You-Shyang Chen;Ching-Hsue Cheng;Chien-Jung Lai;Cheng-Yi Hsu;Han-Jhou Syu

  • Affiliations:

  • Department of Information Management, Hwa Hsia Institute of Technology, 111, Gong Jhuan Rd., Chung Ho District, New Taipei City 235, Taiwan;Department of Information Management, National Yunlin University of Science and Technology, 123, Section 3, University Road, Touliu, Yunlin 640, Taiwan;Department of Distribution Management, National Chin-Yi University of Technology, 35, Lane 215, Section 1, Chung-Shan Road, Taiping District, Taichung 411, Taiwan;Lin's ENT Clinic 364, Fuxing Rd., Lugang Township, Changhua County 505, Taiwan;Department of Information Management, National Yunlin University of Science and Technology, 123, Section 3, University Road, Touliu, Yunlin 640, Taiwan

  • Venue:
  • Computers in Biology and Medicine
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Identifying patients in a Target Customer Segment (TCS) is important to determine the demand for, and to appropriately allocate resources for, health care services. The purpose of this study is to propose a two-stage clustering-classification model through (1) initially integrating the RFM attribute and K-means algorithm for clustering the TCS patients and (2) then integrating the global discretization method and the rough set theory for classifying hospitalized departments and optimizing health care services. To assess the performance of the proposed model, a dataset was used from a representative hospital (termed Hospital-A) that was extracted from a database from an empirical study in Taiwan comprised of 183,947 samples that were characterized by 44 attributes during 2008. The proposed model was compared with three techniques, Decision Tree, Naive Bayes, and Multilayer Perceptron, and the empirical results showed significant promise of its accuracy. The generated knowledge-based rules provide useful information to maximize resource utilization and support the development of a strategy for decision-making in hospitals. From the findings, 75 patients in the TCS, three hospital departments, and specific diagnostic items were discovered in the data for Hospital-A. A potential determinant for gender differences was found, and the age attribute was not significant to the hospital departments.