Using Markov chain models to predict long-term behaviors of customers

• Main Authors: Chiu-Fen Huang, 黃秋芬
• Other Authors: Hsin-Hung Wu
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/28975998591363779625

碩士 === 國立彰化師範大學 === 企業管理學系 === 100 === Due to the ever-changing society, business environment is very different from the past. Companies can easily find that changes in consumers’ demand are faster than before, and it becomes mores difficult to predict customers’ choices. In addition, understanding the choices and behaviors of customers is critically important nowadays. This study uses Markov chain models to observe customers’ behaviors in the long-term perspectives and uses quality function deployment to combine Markov chain models along with PageRank and Pseudocount to show how the proposed approaches work when the sample of size is small and the transition matrix in irregular. By applying PageRank and Pseudocount, irregular transition matrix can be resolved such that the customers’ behaviors can be analyzed.