TOP-10 DATA MINING CASE STUDIES

摘要: We report on the panel discussion held at ICDM'10 conference top 10 data mining case studies in order to provide a snapshot of where and how techniques have made significant real-world impact. The tasks covered by range from detection anomalies such as cancer, fraud, system failures optimization organizational operations, include automated extraction information unstructured sources. From cases we find that supervised methods prevail while unsupervised play supporting role. Further, domain knowledge is generally required achieve completed solution. Finally, successful applications are more commonly associated with continual improvement rather than single "aha moments" ("nugget") discovery.