Outlier Detection Algorithm Suit for Weka

Abstract

WEKA (Waikato Environment for Knowledge Analysis) is open source which is a collection of machine learning algorithms for data mining tasks. WEKA contains tools for data pre-processing, classification, regression, clustering, association rules, and visualization. The workbench also provides a graphical user interface for easy access to these functionalities. There are several advantages to this software that includes probability and usability. Many researchers in industry and academia including students use this software because of it support the standard data mining tasks with good usability .Also, one of the best features of this software is that, it is well suited for developing new machine learning schemes. Taking advantage of this feature, I extended this software to contain in the outlier detection module as well .I have developed set of outlier algorithms and incorporated them under a new module in the weka software. Currently the weka software includes different interfaces like Simple Command Line interface, Explorer, Experimenter and knowledge flow. Explorer is the main interface where the same functionality can also be obtained through the command line or knowledge flow. Experimenter is used for comparisons on different types of panels which provide access to main component of this workbench.