M215-Survival-Analysis (Fall2018)

For UCLA BIOSTAT M215 (Survival Analysis), which covered statistical methods for analysis of survival data.
Textbook: SURVIVAL ANALYSIS Techniques for Censored and Truncated Data (Klein and Moeschberger )

• Most of homework problems are based on the the textbook (Klein and Moeschberger).

Homework 2 - chapter 4

• Fit and plot the Exponential model
• Estimate the survival functions
• Kaplan-Meier estimator and Nelson-Aalen estimator

Homework 3 - chapter 4

• Left truncation, right censored survival data
• Interval estimation for mean/median survival time
• Confidence bands for the survival function

Homework 4 - chapter 7

• One-sample, log-rank test

Homework 6 - chapter 8,9

• Cox’s proportional hazards model
• Hypothesis testing
• Baseline hazard and estimation of survival function
• Stratified Cox’s model

Homework 7 - chapter 11, 12

• Weibull regression model
• Graphical diagnostics
• Martingale residual plot
• Cox-Snell residual plot
• Anderson plots

Final Project

Survival analysis for Echocardiogram dataset

• Objectives and approaches:

1. Explore the distributions of the covariates and correlations between them
2. Estimate survival function using appropriate variable selection
3. Compute Kaplan-Meier curves for the full set
4. Compare survival function among different pericardial effusion groups.
5. Find an appropriate model to predict the risk of death using risk factors in dataset.
6. Fit a Cox’s proportional hazard model to look at the relationship between the patients
   and explanatory variables and deal with any variables with non-proportional hazards