Comparing the Performance of Weibull, Log-Logistics and Gompertz Survival Models on Oncological Data

Farouq Ndamadu Musa, Aliyu Usman, and Abdulhameed Amoto

Abstract

Cancer is a general term used for a group of diseases that cause abnormal cells to divide without control and overpass other tissues. In addition, if they expand out of control, cancer can result in death, an estimated 14.1 million new cases of cancer and 8.2 million deaths from cancer occurred in 2012 in both sexes. The general aim of the research was to compare the performance of Weibull, Log-logistics and Gompertz survival models on oncological data. The research methodology adopted cases on oncological study was used in this study, obtained from internet sources and publications. Descriptive Statistics of dataset was performed using mean, median, mode, variance, skewness, and kurtosis. Parametric survival models were used in the analysis. The models of Weibull, Log-logistics and Gompertz models were chosen because of their similarities in order to have better basis for comparison and also have differences that will cater to the situation where the other one fails. The models are to be fitted to the data with the view to find the best fit, R statistical package was used in analyzing the data. The result revealed that the maximum likelihood estimates of dataset2 with 3 with different model fit, all the information criteria and log-likelihood of the models, Gompertz model has smallest value in all the information criteria, which indicates that Gompertz model is the best fitted model to Myelogenous leukemia data. It also shows the maximum likelihood estimates of dataset 1 with 3 with different model fit, all the information criteria and log-likelihood of the models. Weibull and Log-logistic models have smallest value in all the information criteria, indicating that Weibull and log-logistic models perform better than Gompertz model in fatigue fracture data. The research concludes that Gompertz model is the best fitted model to Remission Times of Bladder Cancer patients’ data, and Gompertz distribution is the best fit distribution for the data and that Gompertz model and Gompertz distribution is also the best fit to Myelogenous leukemia data. It recommended that Gompertz model is the best fit in the oncological data, followed by the Log-logistic model, Weibull and Log-logistic model behave similarly on the dataset.  

Keywords

data; cancer; performance; models; comparison and similarities

Cite This Article

Musa, F. N., Usman, A., Amoto, A. (2024). Comparing the Performance of Weibull, Log-Logistics and Gompertz Survival Models on Oncological Data. International Journal of Scientific Advances (IJSCIA), Volume 5| Issue 3: May-Jun 2024, Pages 428-436, URL: https://www.ijscia.com/wp-content/uploads/2024/05/Volume5-Issue3-May-Jun-No.604-428-436.pdf

Volume 5 | Issue 3: May-Jun 2024