Butterworth Environmental Dataset: Employing COVRATIO for Outlier Detection with Simultaneous Linear Functional Relationship Model

• Published in: IOP Conference Series: Earth and Environmental Science
• Main Author: Al-Hameefatul Jamaliyatul N.A.; Mokhtar N.A.; Badyalina B.; Rambli A.; Zubairi Y.Z.
• Format: Conference paper
• Language: English
• Published: Institute of Physics 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85214849348&doi=10.1088%2f1755-1315%2f1432%2f1%2f012004&partnerID=40&md5=ce6f8a0dccfa458d0e92e2439644d992

The monsoon is affected by factors like wind speed, humidity, and temperature. Outliers can signiicantly skew the data, and this study presents an outlier detection method using the COVRATIO statistic, derived from the covariance matrix within a simultaneous Linear Functional Relationship Model (LFRM) for linear variables. The cut-off point for the 5% upper percentiles of the maximum value of the COVRATIO statistic is established through a Monte Carlo simulation study. The indings indicate that outliers are detected when the COVRATIO statistic surpasses these cut-off points. The effectiveness of the simultaneous LFRM is demonstrated using Butterworth environmental data, with variables including wind speed, humidity, and temperature. The data's normality is conirmed by the Kolmogorov-Smirnov test. This research supports the National Policy on Climate Change by contributing to knowledge-based decision-making in climate-related studies, particularly in the domains of environmental monitoring, renewable energy planning, and data analysis. © 2024 Institute of Physics Publishing. All rights reserved.