Outlier detection strategy for distributed high dimensional data sets with mixed attributes projects main idea is to implement new methods in dataset which is mostly used in research fields of credit card fraud detection and other fraud cases in electronic commerce. This paper will explain different methods for finding patterns that mostly occur in the dataset and compare it with existing data mining techniques which are used to find regular patterns.
In existing research papers outlier detection is mostly focused on datasets with a specific attribute type, by considering these attributes as numerical and ordinal or categorical. Among these methods categorical is not a accurate process where categorical data is converted to numerical values.
download detecting outlier detection strategy for distributed high dimensional data sets with mixed attributes project reference document.