Exceptional Patterns with Clustering Items in Multiple Databases

摘要

The information from multiple local databases can be mined together to make global patterns. More global decisions will be based on synthesized patterns and exceptional patterns using clustering technique. To take better decision in organization head quarter level, exceptional patterns also need to be analyzed for non profitable things which help for continuous company growth. Our new strategy is developed considering both clustering the frequent items also exceptional patterns. Various experiments are conducted with 10 sample datasets and the results are recorded in experimental section, with this the significance and limitations behind those approaches can be made clear.