Local outlier detection by integrating fuzzy constraint and ensemble framework using Bio inspired algorithms

Abstract

The main objective of the work is to detect local outliers by newlineintegrating the fuzzy constraint and ensemble framework using Bio-Inspired newlineAlgorithms. Outliers are the data values that deviate significantly from the newlinemajority of the data and their values normally fall aside from the overall newlinepattern of the data. Outlier detection aims to determine such type of dissimilar newlineand exceptional events. newlineThe thesis is divided into six chapters. Chapter one gives an newlineoverview of introduction to outlier detection. Outlier detection finds newlinewidespread applications in intrusion detection, fraud detection, medical newlinediagnosis, fake news detection, sensor networks and so on. The different types newlineof outliers, classification of outlier detection methods based on data labels and newlinemethodologies, challenges in outlier detection are presented in detail. newlineThe motivation of the research and the research objectives are also newlinehighlighted in this section. newlineChapter two provides a detailed review of literature pertaining to newlinethe proposed work. The survey includes the methodologies, advantages and newlinedrawbacks of the outlier detection methods, subspace methods, fuzzy newlineapproaches, ensemble based methods and bio inspired algorithms. newlineThis section also outlines the outlier detection methods on high dimensional newlinedata, mixed type data, categorical data and multi view data. newline