International Journal of Computer Application and Engineering Technology (IJCAET)
Outlier detection is concerned with discovering the exceptional behaviors of certain objects. The process of finding, removing and detecting outliers is a complex phenomena in data mining .The data stored in large database are prone to errors is property of a data mining algorithm with respect to outliers in the database. Many data mining methods in data mining address this problem to some extent, but not fully the problem is solved. The authors present a Scaled Outlier K-means Algorithm (SOKA) algorithm that provides outlier detection and removal on multidimensional data. They also provide experimental results that show that this is indeed a practical solution to the multidimensional outliers.