Classification of data mining techniques in intrusion detection

Gaurav Mishra

ISBN 10: 3659442151 / ISBN 13: 9783659442155

Published by LAP LAMBERT Academic Publishing Aug 2013, 2013

About this Item

Neuware -Data mining is the process of gathering, searching, and analyzing a large amount of raw data, as to discover patterns, relationships and behavior of data. There are large numbers of algorithms for classification of data mining. Single algorithm is not efficient for classification of data and recognize their pattern and behavior .There is a key term known as ensemble learning which means Combining two or more classifiers for efficient result. I have used the KDD¿99 dataset for the experiment which have 41 features labeled either as normal or as an attack. In this book I have represented how graphical machine learning tool weka can be used for data mining and how ensemble learning can be implemented using weka.I have used three classifiers with the Bagging and Boosting ensemble learning approach which are complementary naïve bayes and two are rule based classifiers, part and jrip. My experiment shows that bagging improves the efficiency of the rule based classifiers as well as of naïve Bayes. However, the rule based classifiers become more efficient with bagging and boosting techniques.Books on Demand GmbH, Überseering 33, 22297 Hamburg 52 pp. Englisch.

Seller Inventory # 9783659442155

Bibliographic Details

• Title: Classification of data mining techniques in intrusion detection
• Publisher: LAP LAMBERT Academic Publishing Aug 2013
• Publication Date: 2013
• Language: English
• ISBN 10: 3659442151
• ISBN 13: 9783659442155
• Binding: Taschenbuch
• Condition: Neu
• Dimensions: 220W x 150H x 4D millimeters 220 millimeters width by 150 millimeters height by 4 millimeters depth
• Weight: 96 grams

About this title

Synopsis

Data mining is the process of gathering, searching, and analyzing a large amount of raw data, as to discover patterns, relationships and behavior of data. There are large numbers of algorithms for classification of data mining. Single algorithm is not efficient for classification of data and recognize their pattern and behavior .There is a key term known as ensemble learning which means Combining two or more classifiers for efficient result. I have used the KDD’99 dataset for the experiment which have 41 features labeled either as normal or as an attack. In this book I have represented how graphical machine learning tool weka can be used for data mining and how ensemble learning can be implemented using weka.I have used three classifiers with the Bagging and Boosting ensemble learning approach which are complementary naïve bayes and two are rule based classifiers, part and jrip. My experiment shows that bagging improves the efficiency of the rule based classifiers as well as of naïve Bayes. However, the rule based classifiers become more efficient with bagging and boosting techniques.

About the Author

Gaurav Mishra is born in India and has received His Bachelor’s degree in Computer Science & Engg.From Malout Institute of Management & Information Technology, Malout India. His areas of interest are Network Security, Java programming & Data Mining.