Interval-estimation of Customer Equity Based on Product-customer Classification

摘要

The paper applies clustering method to classify the customers according to their purchase records on the basis of classification of product. Then it explores the change rules of the amount and expenditure of every category of customers by virtue of regression analysis. Interval-estimation of customer equity is developed after the two steps. The estimation method can accurately predict customers' future expenditure by classification, and interval-estimation may reflect that customer equity is a random variable, thus to offer high quality data as a reference to marketing decisions.