## 2017 Graduate Online Symposium

### Survival Analysis of Breast Cancer Using Parametric Methods

Event

#### Description

Survival analysis involves the study of time until an event of interest occurs, or lifetimes. This analysis is used to analyze how long after patients are diagnosed with breast cancer they can live while studying the variables or covariates involved. Hypotheses tests are performed on the different attributes and conclusions are made based on their significance in predicting survival. In this research, I shall perform hypothesis testing to determine differences in survival functions for different groups.

Parametric models are used to represent the relationship between the survival time in months of patients diagnosed with breast cancer and the covariates or the explanatory variables. The parametric models we focus on in this study include Weibull distribution, Exponential distribution, Log-normal, and log- Logistic. These distribution models are thoroughly explained and analyzed for proportional hazards in this study. Accelerated failure time (AFT) models which are parametric regression analysis for log- linear models is discussed and a comparison is made to the parametric proportional hazard function.

#### Keywords

survival, breast cancer, ethnicity, parametric

#### Degree

Master of Science (MS)

#### Department

Mathematics and Statistics

#### College

Science, Engineering and Technology

Mezbahur Rahman

#### First Faculty Advisor's Department

Mathematics and Statistics

#### First Faculty Advisor's College

Science, Engineering and Technology

#### Creative Commons License

Jepng'etich_Lizzy_Handout.pdf (586 kB)
Presentation Handout

#### Share

COinS

Apr 17th, 12:00 AM Apr 17th, 12:00 AM

Survival Analysis of Breast Cancer Using Parametric Methods

Survival analysis involves the study of time until an event of interest occurs, or lifetimes. This analysis is used to analyze how long after patients are diagnosed with breast cancer they can live while studying the variables or covariates involved. Hypotheses tests are performed on the different attributes and conclusions are made based on their significance in predicting survival. In this research, I shall perform hypothesis testing to determine differences in survival functions for different groups.

Parametric models are used to represent the relationship between the survival time in months of patients diagnosed with breast cancer and the covariates or the explanatory variables. The parametric models we focus on in this study include Weibull distribution, Exponential distribution, Log-normal, and log- Logistic. These distribution models are thoroughly explained and analyzed for proportional hazards in this study. Accelerated failure time (AFT) models which are parametric regression analysis for log- linear models is discussed and a comparison is made to the parametric proportional hazard function.