Analysis & implementation of item based collaboration filtering using K-Medoid

This thesis uses data mining classification algorithm classification algorithms to get useful information to decision-making out of customer ship transaction behaviors. Firstly, by business understanding, data understanding and data preparing, modeling and evaluating we get the results of the two algorithms and by comparing the results, we know that the two algorithms can both be applied in the customer membership card classification model and can obtain a quite accurate result. Then we introduce the application of this model. In classification tree modeling the data is classified to make predictions about new data. Using old data to predict new data has the danger of being too fitted on the old data. But that problem can be solved by pruning methods which decentralizes the modeled tree. This paper describes the use of classification trees and shows two methods of pruning them. An experiment has been set up using different kinds of classification tree algorithms with different pruning methods to test the performance of the algorithms and Pruning methods.