Proposal Description

Analysis of lifetime data is very important in many disciplines including Survival and Reliability Analysis. It is also important to know the effect of covariates on the lifetime of a subject. The proportional hazards model and proportional odds model are well discussed in the right censored data to incorporate the covariate effect on Survival times. Different types of censoring schemes such as current status censoring, double censoring and interval censoring are common in survival studies. In this proposal, we proposed regression models for analysing different types of survival data under different censoring schemes.
Competing risks emerge naturally in lifetime data analysis when the subjects under study are at risk for more than one cause of failure. For example, consider a study of patients suffering from heart disease. The patients may die due to other causes like accidents or other diseases, which may alter the probability of death due to heart disease. Hence it is important to analysis the survival data by considering causes and covariates.
In many longitudinal studies on recurrent events, instead of observing the exact time of occurrence of an event, we may only observe the number of recurrences experienced by a subject in a given period. If each subject can be observed at more than one-time point, the number of recurrences between two successive observation times is available. The data obtained in this form is known as panel count data. We develop a semiparametric regression models model for panel count data when the study subjects are exposed to recurrent events due to more than one mode of recurrence. We proposed a regression model under this setup
The current status censoring takes place in survival analysis when the exact event times are not known, but each individual is monitored once for their survival status. The current status data often arise in medical research, from situations that involve multiple causes of failure. Examining current status competing risks data, commonly encountered in epidemiological studies and clinical trials, is more advantageous with Bayesian methods compared to conventional approaches. They excel in integrating prior knowledge with the observed data and delivering accurate results even with small samples. Inspired by these advantages, the present study is pioneering in introducing a Bayesian framework for both modeling and analysis of current status competing risks data together with covariates.
In our study, we assume independence censoring. That is, given covariates, the lifetime and censoring time are independent. It is desirable to develop a regression model under dependence censoring. In recent times, Machine learning methods in survival analysis have become very popular. We combine the machine learning and survival model to address the problem of dependent censoring.