Robust methods for detecting bad leverage point in logistic regression

Year 2024, Volume: 42 Issue: 5, 1344 - 1356, 04.10.2024

Ebru Gündoğan Aşık , Arzu Altin Yavuz , Zafer Küçük

Abstract

High-leverage points, known as good and bad leverage points, are also known as points away from center of x space. Bad leverage points are marginal values that show the incompatibility with misclassified observations and other observation values at x space. In the identification of bad leverage points, the problems of masking and swamping constitute a problem for the logistic regression model just as in the linear regression model. In this research, in addition to existing deviance components (DEVC), robust deviance components (RobDEVC) that are used to identify bad leverage points, different robust methods recommended to be used at the management of deviance components were examined. Also, for these methods, robust cut-off value combinations were examined as well. With the conducted simulation, robust methods recommended to be used in the deviance component method have shown better performance to identify bad leverage points by showing different cut-off values.

Keywords

Bad Leverage, High-Leverage, Logistic Regression, Robust Cut-Off Value, Robust Deviance Residuals

References

• REFERENCES
• [1] Belsley DA, Kuh E, Welsch RE. Regression Diagnostics. Hoboken, NJ, USA: John Wiley & Sons, Inc.; 1980. [CrossRef]
• [2] Imon AHM. Identification of high leverage points in logistic regression. Pakistan J Stat 2006;22:147–156.
• [3] Norazan MR, Sanizah A, Habshah M. Identifying bad leverage points in logistic regression model based on robust deviance components. Math Models Methods Modern Sci 2012:62–67. [CrossRef]
• [4] Nurunnabi AAM, Imon AHMR, Nasser M. Identification of multiple influential observations in logistic regression. J Appl Stat 2010;37:1605–1624. [CrossRef]
• [5] Syaiba BA, Habshah M. Robust logistic diagnostic for the identification of high leverage points in logistic regression model. J Appl Sci 2010;10:3042–3050. [CrossRef]
• [6] Arqub OA, Abo-Hammour Z. Numerical solution of systems of second-order boundary value problems using continuous genetic algorithm. Inform Sci 2014;279:396–415. [CrossRef]
• [7] Abu Arqub O, Singh J, Alhodaly M. Adaptation of kernel functions-based approach with Atangana–Baleanu–Caputo distributed order derivative for solutions of fuzzy fractional Volterra and Fredholm integrodifferential equations. Math Methods Appl Sci 2021:7807–7834. [CrossRef]
• [8] Abo-Hammour Z, Abu Arqub O, Momani S, Shawagfeh N. Optimization solution of Troesch’s and Bratu’s problems of ordinary type using novel continuous genetic algorithm. Discrete Dyn Nat Soc 2014;2014. [CrossRef]
• [9] Ahmad S, Ramli NM, Midi H. Robust estimators in logistic regression: A comparative simulation study. J Modern Appl Stat Methods 2010;9:502–511. [CrossRef]
• [10] Croux C, Haesbroeck G. Implementing the bianco and yohai estimator for logistic regression. Comput Stat Data Anal 2003;44:273–295. [CrossRef]]
• [11] Venter JH, De La Rey T. Detecting outliers using weights. South African Stat J 2007;41:127–160.
• [12] Gündoğan Aşik E, Altin Yavuz A, Küçük Z. New robust cut-off values in determining bad leverage points in the logistic regression model. Mehmet Akif Ersoy Univ J Econ Administrat Sci Fac 2021;8:630–650. [Turkish] [CrossRef]
• [13] Sarkar K, Midi H, Rana S. Detection of outliers and influential observations in binary Logistic regression: An empirical study. J Appl Sci 2011;11:26–35. [CrossRef]
• [14] Künsch HR, Stefanski LA, Carroll RJ. Conditionally unbiased bounded-influence estimation in general regression models, with applications to generalized linear models. J Am Stat Assoc 1989;84:460–466. [CrossRef]
• [15] Carroll RJ, Pederson S. On robustness in the logistic regression model. J R Stat Soc Series B Stat Methodol 1993;55:693–706. [CrossRef]
• [16] Pregibon D. Logistic regression diagnostics. Ann Stat 1981;9:705–724. [CrossRef]
• [17] Bianco AM, Yohai VJ. Robust Estimation in the Logistic Regression Model. Springer V. New York, NY: 1996. [CrossRef]
• [18] Croux C, Haesbroeck G. Implementing the Bianco and Yohai estimator for logistic regression. Comput Stat Data Anal 2003;44:273–295. [CrossRef]
• [19] Rousseeuw PJ, Leroy AM. Robust Regression and Outlier Detection. Hoboken, NJ, USA: John Wiley & Sons, Inc.; 1987. [CrossRef]
• [20] Hastings C, Mosteller F, Tukey JW, Winsor CP. Low moments for small samples: A comparative study of order statistics. Ann Math Stat 1947;18:413–426. [CrossRef]
• [21] Hoaglin DC, Mosteller F. Understanding robust and explonatory data analysis,. New York, NY: John Wiley & Sons; 1983.
• [22] Hodges JLJ, Lehmann EL. Estimates of location based on ranks tests. Ann Math Stat 1963;34:598–611. [CrossRef]
• [23] Nurunnabi AAM, Imon AHMR, Nasser M. A new statistic for influence in linear regression. Technometrics 2008;47:1–11.