Segmented Quantile Regressions based on the Unit-Weibull Distribution for bounded data

Shao-Tung Chang () and Kang-Ping Lu ()
Additional contact information
Shao-Tung Chang: Department of Mathematics, National Taiwan Normal University Taipei
Kang-Ping Lu: Department of Applied Statistics, National Taichung University of Science and Technology Taichung

No 15216620, Proceedings of International Academic Conferences from International Institute of Social and Economic Sciences

Abstract: In this study, we introduce a segmented quantile regression model based on the unit-Weibull distribution. Given that the unit-Weibull random variable is defined on the (0,1) interval, it is particularly suitable for analyzing continuous variables constrained within this range. Such variables, including fractions, indices, ratios, and percentages, are prevalent across various disciplines. Additionally, many bounded variables can be transformed into unit-interval data, further expanding the utility of the unit-Weibull distribution.The segmented regression model is a widely used technique for identifying structural changes in regression models by detecting turning points, which can provide insights into underlying causal factors. Moreover, compared to ordinary least squares regressions, quantile regressions offer greater robustness against outliers and non-normal error distributions. We incorporate the unit-Weibull distribution with segmented and quantile regressions to develop a flexible and resilient modeling approach applicable across diverse research domains. By using the maximum likelihood estimation with the expectation and maximization algorithm, we investigate model estimations as well as applicable data. This research is novel, highly practical, versatile, relatively easy to use. Besides, we reduce the restriction in using the segmented regression to enlarge the applicability of our method in versatile data and research fields. We demonstrate the efficiency and robustness of our new methods through extensive simulations and experiments. The practicability of the proposed methods is shown by applying the method to various real datasets.

Keywords: Quantile regressions; Unit-Weibull distribution; Segmented regressions; Segmented quantile Regression models (search for similar items in EconPapers)
JEL-codes: C01 C10 C19 (search for similar items in EconPapers)
Pages: 1 page
References: Add references at CitEc
Citations:

Published in Proceedings of the Proceedings of the 66th International Academic Conference, London, Nov -0001, pages 9-9

Downloads: (external link)
https://iises.net/proceedings/iises-international- ... 52&iid=003&rid=16620 First version, 0000

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:sek:iacpro:15216620

Access Statistics for this paper

More papers in Proceedings of International Academic Conferences from International Institute of Social and Economic Sciences
Bibliographic data for series maintained by Klara Cermakova ().