Course overview of AST405
(AST405) Lifetime data analysis
Instructor
Md Rasel Biswas
MS & BS in Applied Statistics (DU)
Lecturer at DU since July 2022
Email: rbiswas1@isrt.ac.bd
Course introduction
Course Title: Lifetime Data Analysis
Credit Hour: 4
This course focuses on the analysis of survival or failure-time data, widely used in fields such as medicine, clinical trials, biology, epidemiology, engineering, finance, and social sciences. Students will explore lifetime probability distributions, non-parametric methods, parametric models and estimation, accelerated failure time models, and proportional hazard models for analyzing lifetime data.
Course objectives
Upon completing this course, students will be able to:
- Identify and analyze censored data using appropriate statistical methods and models.
- Understand and apply the theory and methodology for analyzing lifetime data from both complete and censored samples.
- Interpret statistical lifetime distributions, types of censoring, and utilize graphical techniques for data exploration.
- Implement non-parametric and parametric estimation methods in survival analysis.
- Apply lifetime regression models, including accelerated failure time and proportional hazards models, in practical scenarios.
Lecture plans
Lecture 1- 6: Basic concepts and models: lifetime distributions-continuous models, discrete models, a general formulation; some important models-exponential, Weibull, log-normal, log-logistic, gamma distributions, log-location-scale models, inverse Gaussian distributions models, mixture; regression models.
Lecture 7- 10: Observation schemes, censoring, and likelihood: right censoring and maximum likelihood; other forms of incomplete data; truncation and selection effects; information and design issues.
Lecture 11- 17: Nonparametric and graphical procedures: nonparametric estimation of survivor function and quantiles; descriptive and diagnostic plots; estimation of hazard or density functions; methods of truncated and interval censored data; life tables.
Lecture 18- 24: Inference procedures for parametric models: inference procedures for exponential distributions; gamma distributions; inverse Gaussian distributions; grouped, interval censored, or truncated data; mixture models; threshold parameters; prediction intervals.
Lecture 25- 30: Inference procedure for log-location-scale distributions: inference for location-scale distributions; Weibull and extreme-value distributions; log-normal and log-logistic distributions; comparison of distributions; models with additional shape parameters; planning experiment for life tests.
Lecture 31- 36: Parametric regression models: introduction to log-location-scale regression models, proportional hazards regression models; graphical methods and model assessment; inference for log-location-scale models; extensions of log-location-scale models; hazard based models.
Lecture 37- 40: Brief introduction to Cox’s proportional hazards model; partial likelihood function, estimation and interpretation of model parameters.
Textbooks
Lecture time
Every Monday and Wednesday 11:20 AM – 12:50 PM
Assessment
- Attendance: 5%
- Incourse exams: 25%
- Final exam: 70%