Wednesday 20th of September 2017
 

A Model to Find Outliers in Mixed-Attribute Data sets using Mixed Attribute Outlier Factor


M. Krishna Murthy, A. Govardhan and Lakshmi Sreenivasareddy D

Outliers are records in real datasets which have abnormal behavior comparing with other records in datasets. Finding outliers in numerical dataset is easy. Many methods are available for numerical datasets. Number of methods is also available for categorical datasets. But very less number of methods are available for mixed attribute Datasets. In available methods the concept of frequent pattern mining is used. Finding different frequent patterns from datasets for the categorical attributes is a cumbersome process. In this proposed model Mixed Attribute Outlier Factor (MAOF) is presented. Which is a simple technique and it requires only one scan of data set. MAOF is derived based on Attribute Value frequency for Categorical part of data set and cosine factor from mean value order set to remaining numerical data points in the data set. Average of these two factors will give the MAOF score. This model is applied on a Bank data set which is a real data set taken from UCI ML repository [10]. This method shows the good results.

Keywords: Data mining, Outlier detection, Otey score, ODMAD score, MAOF score

Download Full-Text


ABOUT THE AUTHORS

M. Krishna Murthy
Mr. Mudimbi.Krishna Murthy did His M.C.A in first Class in 2003 from MKU Madurai. He has 15 years of technical experience in Computer Science and Engineering at School of Information Technology (SIT), Jawaharlal Nehru Technological University Hyderabad, India. He has five research papers at international and national conferences. His area of research is Data Mining and Information Retrieval Systems.

A. Govardhan
Dr.A.Govardhan is presently a Professor of Computer Science & Engineering, Jawaharlal Nehru Technological University Hyderabad (JNTUH), India. He did his B.E (CSE) from Osmania University College of Engineering, Hyderabad in 1992, M.Tech from Jawaharlal Nehru University (JNU), New Delhi in 1994 and Ph.D from Jawaharlal Nehru Technological University, Hyderabad in 2003. He is a recipient of several International and National Awards including A.P. State Best Teacher Award, Bharat Seva Ratna Puraskar, CSI Chapter Patron Award, Bharat Jyoti Award and Mother Teresa Award for Outstanding Services, Achievements, Contributions for Meritorious Services, Outstanding Performance and Remarkable Role in the field of Education and Service to the Nation. He is a Chairman and Member on several Boards of Studies of various Universities. He is the Chairman of CSI Hyderabad Chapter. He is a Member on the Editorial Boards for Eight International Journals. He is Member of several Advisory Boards and Committee Member for several International and National Conferences. He has guided 15 Ph.D theses and he has published 152 papers at International/National Journals/Conferences including IEEE, ACM, Springer and Elsevier. He has delivered more than 35 Keynote addresses and invited lectures. He served as Principal, Head of the Department and Studentsâ Advisor. He is a member in several Professional and Service oriented bodies. His areas of research include Databases, Data Warehousing & Mining and Information Retrieval Systems.

Lakshmi Sreenivasareddy D
Mr.LakshmiSreenivasareddy.D obtained his Masters degree from Jawaharlal Nehru Technological University Hyderabad (JNTU). He is pursuing his Ph.D in Computer Science and Engineering from JNTUH, Hyderabad. He is currently heading the Department of Computer Science & Engineering, RISE Gandhi Groups of Institutions Ongole. He has 10 years of teaching experience. He has six research papers at international conferences and journals including IEEE and Elsevier. His area of interest is Data Warehousing & Mining and Information Retrieval Systems.


IJCSI Published Papers Indexed By:

 

 

 

 
About IJCSI

IJCSI is a refereed open access international journal for scientific papers dealing in all areas of computer science research...

Learn more »
Join Us
FAQs

Read the most frequently asked questions about IJCSI.

Frequently Asked Questions (FAQs) »
Get in touch

Phone: +230 911 5482
Email: info@ijcsi.org

More contact details »