Estimation of Baseline Hazard Function on Cox Model using Maximum Penalized Likelihood

  • Almira Ivah Edina, Santi Wulan Puranami, Purhadi, Edi Sukur

Abstract

In Cox regression, the baseline hazard function is unknown. Baseline hazard function
is a part of hazard function which can be used for modeling. Without baseline hazard
function, there will be important components missing and the result cannot be used to
report the whole model. This research review maximum penalized likelihood (MPL) to
estimate the baseline hazard because this method has a small bias when compared to the
partial likelihood method and applied this method on brain cancer patients at C-Tech
Labs Edwar Technology, Tangerang. From the result, we know that the more often brain
cancer patients consul, they have a smaller risk of dying. In addition, it is also known that
brain cancer patients who do radiotherapy have a risk of dying 2.578959 times greater
than those who do not do radiotherapy. We also know that the death rate of brain cancer
patients from day 0 until day 500 (2 years) up and down quickly, but after day 500 (2
years) until day 1500 (4 years) the death rate up and down slowly. Where, the day after
day 1500 (4 years) the death rate of brain cancer patients is constant.

Published
2020-05-01
How to Cite
Almira Ivah Edina, Santi Wulan Puranami, Purhadi, Edi Sukur. (2020). Estimation of Baseline Hazard Function on Cox Model using Maximum Penalized Likelihood. International Journal of Advanced Science and Technology, 29(7s), 3253-3263. Retrieved from https://sersc.org/journals/index.php/IJAST/article/view/17606