TCSOM: Clustering Transactions Using Self-Organizing Map
Neural Processing Letters
k-ANMI: A mutual information based clustering algorithm for categorical data
Information Fusion
Hi-index | 0.00 |
Transaction clustering has received attentions in recentdevelopments of data mining.Traditional clustering methodsare not useful to solve this problem.Transaction datasets are different from the traditional data sets in their highdimensionality, sparsity and numerous outliers.In this paper,we introduce a new efficient algorithm for transactionclustering.The proposal algorithm is based on a caucus,which fine-partitioned demographic groups based on purchasefeatures of customers.Due to the important role caucusplays, we also present a heuristic method of caucus generationwith the use of entropy.Experiments on real and synthetic data sets show that our approach can achieve a better result than existed methods.