30 October، 2022
Discussion of a master’s thesis in the College of Computer Science and Mathematics – Department of Statistics and Informatics entitled: The Use of Robust Weight Cox Regression with Bootstrap in Analyzing Data for Women with Breast Cancer in Nineveh Province
Discussion of a master’s thesis in the College of Computer Science and Mathematics – Department of Statistics and Informatics entitled:The Use of Robust Weight Cox Regression with Bootstrap in Analyzing Data for Women with Breast Cancer in Nineveh ProvinceIt was discussed in the discussion room at the Faculty of Computer Science and Mathematics at the University of Mosul on Sunday 30/10/2022Master’s thesis by student Salwa Salah ALdeen Qassim Haidari, under the supervision of Assist. Prof.Dr. Bashar A. Majeed AL-TalibThis focuses on the estimation of survival time for real data representing breast cancer patients’ data in Nineveh Governorate for the period from 2007 to 2013, and robust formulas will be used to estimate and deal with the Cox regression model in survival analysis and the formation of life tables for patients and to measure the most important factors affecting the survival time of people afflicted with the disease. It will suggest the best Robustness model to be used in determining the degree of Hazard facing patients; The thesis seeks to suggest the use of some Robust weights and the replacement of traditional variance estimators with robust estimations and experimenting with a set of alternatives to reach an efficient estimate of the model with some comparison criteria after simulating the estimated model using the bootstrap technique to generate samples from the real data collected.The study touched the Survival analysis, which is an important field of biomedical research, especially research of incurable diseases such as cancer, and the need has arisen to develop statistical methods in estimating models to increase accuracy and knowledge of the factors that will affect the patient’s survival by adapting the methods used to confront any irregularities in the assumptions of the analysis The model and these methods or means have regression models, especially models that fit the case of the dependent variable, which will be binary in response, as is the case in the data of cancer patients (a time-related binary variable). One of these models is the Cox regression model proposed by the scientist (Cox) in 1972 and that the formula The Cox regression model is as followsSince are independent variables and β is the model coefficients and is the basic hazard function, we have resorted to applying robust methods in estimating it to face the possibility of abnormal values in the data whose main objective is to estimate parameters A model that represents information about the majority of the data;This study aims to :
- Estimation of Cox’s Robust regression model using Pre-Defined Templates method and three templates were applied as follows:AHR “Average Hazard Ratios”: This template estimates the average hazard ratios using some weights with a so-called Censoring Correction and a Robust variance estimator. “ARE” Average Regression Effects: This template estimates the average regression effects using a Censoring Correction and a cox proportional hazards regression, as in the previous template. PH “Cox proportional hazards regression”
- Using the traditional (Lin, 1991) method: in it, weight functions usually used in tests of the weighted order logarithm are combined with the partial potential function.
- Using the method (Sasieni, 1993): This method is based on maximizing the partial potential function weighted by Robust weights.
- Applying the generalized Akaki criterion with the three templates (PH, AHR, ARE) to arrive at the best proposed model.
- Using proposed robust weight functions such as Huber’s weight function and applying it with the three templates to arrive at the best model that has an impact on the occurrence of the event.The discussion committee consists of:
Prof. Dr. Zakaria Yahya Nouri – Chairman Assis. Prof. Dr. Dilshad Shaker Ismail/ Salahaddin University – member Assis. Prof. Dr . Osama Bashir Hanoun – member Assis. Prof. Dr . Bashar Abdel Aziz Al-Talib – member and supervisor Contact the university
University Address
Al Majmoaa Street
Postal Code : 41002
Mosul – Iraq
All Rights Reserved for University of Mosul - 2024
Go to Top
