Detection of Outliers Using Hybrid Algorithm on Categorical Datasets

The outlier is an observation that is different from the other remaining values in a data set. Real life contains large number of categorical data. There is some outlier detection algorithms have been designed for categorical data. There are two main problems of outlier detection for categorical data, which are the time complexity and accuracy for detection of outliers in categorical dataset. Categorical dataset have some limited approaches as compared to numeric dataset. This paper describes about some existing algorithms for outlier detection in categorical dataset. The novel Hybrid method which overcomes limitations of previous methods (NAVF and ROAD) has been implemented. The algorithm is implemented and tested on different types of networking datasets, in which detected outliers are virus or intrusion whose behavior is different than behavior in normal networking data.