| 摘要 |
PROBLEM TO BE SOLVED: To provide a method and a device for determining one ore more statistical estimated quantities regarding customer's trends which can overcome or at least reduce one or more problems. SOLUTION: Transaction information on a customer and other customer information are held. Provided is a means which makes estimates regarding future customer's trends, etc., by estimating the possibility that the customer leaves a bank by using the information. A business executive can take measures for improving the result by using those estimates. A Bayes statistic model is generated by using customer data and used to generates a statistical estimated quantity of customer's trends. This statistical model is formed by using the hidden Markov model technique by clustering data and properties of the customer into a finite number of states. The number of the states is not observed and considered to be random. Bayes posterior probability distributions are specified and combined with the data to generate Bayes posterior probability distributions. A statistical estimate can be obtained by using those Bayes posterior probability distributions. |
