Segmentation data exploration methods in modern real-time data warehouse

In modern business intelligence systems there is a need to reduce the data flow between operational transactional systems and data warehouses, as well as reduce the time between data update in the warehouse and reflecting this change in the analytical models used to perform business analyses. In a modern data warehouse with incremental data update, at every change there is only a small amount of new data present, however most of the data mining techniques requires training the model on the full training set. In this paper popular data mining segmentation techniques are presented along with incremental learning algorithms, as well as a new segmentation method with the use of genetic algorithm.