Fuzzy Membership Function: Outlier Detecting Methods and Detecting the Outlier by Using Five Number Summary of Data

  Mounika.P,  Sowbarnikaa.A,  Shamruthi.M

In the fuzzy system describing the membership function is very important activity in data science. An outlier is the thing which is consider extremely dissimilar from the rest of the other objects. Distance based methodology, density based methodology and deviation based methodology and five number summary of data etc., are the methods which plays a vital role in identifying the outliers. Deceptive results in many things is obtained due to the presence of outliers. So it is very important to identify and eliminate the outliers in order to prevent the bad impressions which is created due to the presence of outliers. Then in human error, environmental changes, error in the instrument and in malicious activity etc.; detecting the outliers plays a important role in these applications. In this paper constructing fuzzy membership function by using the five number summary of data method and the other methods which is involved in outlier detection have been discussed and explained respectively. The main advantage is that the accuracy of prediction have been obtained while removing the outliers.

Keywords- Outliers, Fuzzy Set, Membership Function, Five Number Summary data, density based method, deviation based method, distance based method.

Unique Identification Number - IJEDR2004040
Page Number(s) - 251-257
Pubished in - Volume 8 | Issue 4 | December 2020
DOI (Digital Object Identifier) -   
Publisher - IJEDR (ISSN - 2321-9939)

  Mounika.P,  Sowbarnikaa.A,  Shamruthi.M,   "Fuzzy Membership Function: Outlier Detecting Methods and Detecting the Outlier by Using Five Number Summary of Data", International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.8, Issue 4, pp.251-257, December 2020, Available at :http://www.ijedr.org/papers/IJEDR2004040.pdf