UMP Institutional Repository

Detection of different outlier scenarios in circular regression model using single-linkage method

N. M. F., Di and Siti Zanariah, Satari and Roslinazairimah, Zakaria (2017) Detection of different outlier scenarios in circular regression model using single-linkage method. In: Journal of Physics: Conference Series, 1st International Conference on Applied & Industrial Mathematics and Statistics 2017 (ICoAIMS 2017), 8-10 August 2017 , Kuantan, Pahang, Malaysia. pp. 1-5., 890. ISSN 1742-6588 (print); 1742-6596 (online)

[img]
Preview
Pdf
Detection of different outlier scenarios.pdf
Available under License Creative Commons Attribution.

Download (485kB) | Preview

Abstract

Outliers are the set of data that are significantly deviates or dissimilar from the rest of the data set. In circular regression model, the existence of outliers are well known to give a large effect on the parameter estimates and inferences. In this study, we proposed clustering-based method using single linkage to detect multiple outliers. Single-linkage is one of several clustering methods, where the distance between two clusters is determined by a single pair element that are closest to each other. We examined two outlier scenarios with a certain degree of contamination. The performance of proposed method on different outlier scenarios are compared and the best method for each outlier scenario is chosen

Item Type: Conference or Workshop Item (Lecture)
Uncontrolled Keywords: Regression analysis
Subjects: Q Science > QA Mathematics
Faculty/Division: Faculty of Industrial Sciences And Technology
Depositing User: Noorul Farina Arifin
Date Deposited: 03 Oct 2018 04:06
Last Modified: 03 Oct 2018 04:06
URI: http://umpir.ump.edu.my/id/eprint/22307
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item