a new method for missing data imputation in banking datasets based on biclustering

Online banking (internet banking) has emerged as one of the most profitable e-commerce applications over the last decade and thus data analysis and data mining techniques are extensively used to enhance decision making in financial institutions and banks. one of the main challenges in data mining for e-banking is the existence of missing values. a new method is proposed in this paper to impute missing values based on the cross-relationship between information stored in the banking databases. given that all banking information is not stored in a single table and there are useful data in other tables, it is demonstrated how missing values of city attribute in customers table can be estimated using the information stored in transactions table. first a pivot table is generated based on transactions table and then a biclustering algorithm is applied to group customers. finally, missing city values of the customers are imputed using existing ones. the experimental results show that the proposed method has better performance than classic imputation methods and can be easily employed in other similar cases for imputing missing attributes in banking datasets.