Average Clustering of Discrete Data Based on Probability and its Application to Expressway Toll Fraud Detection

This paper introduces a algorithms about average clustering of discrete data based on probability. By using the algorithms we clustered the vehicle plate information in the northern part of Guangdong provincial highway network in 2007 to 35 cluster. We find that different threshold value can produce different result. The cluster method has been calculated and all vehicle plate information at the entrance in the northern part of Guangdong province highway network from 8 o´clock on January 1 of 2008 onwards is put to the database in 35 cluster. This paper also introduces the data in the second quarter of 2008 using the cluster method to cluster the data into the database. It is conclude the produced cluster method can meet the demand of toll collection. This paper also analyses the merits and drawbacks of using the average clustering of discrete data based on probability to cluster the data.