- this was a group project done by a total of two people (Bryan Zang and Linsi Zhong) under the course "Statistical Methods for Life History Analysis" by Professor Liqun Diao
- this project is a novel comparison of existing publications, for educational purposes, all credits go to the rightful owners
- the project is designed to give you some exposure to more advanced and more recently developed statistical or machine learning methods for life history analysis. You are expected to read research papers to learn new method(s) for analyzing longitudinal data or survival data, provide a clear and thorough review of the method(s), implement them and compare them to the methods we learned in the lectures and report your findings.
- team: You may work on this project individually or in a team of two. If you choose to work with a classmate as a team, each team member must contribute equally to the project work.
- topic: You should choose a specific topic related to analysis of life history data (either longitudinal data or survival data). Your project can be based on one or multiple research papers on a common theme. You can choose your paper(s) from the given list of papers on LEARN or outside the list subject to the instructor’s approval.
- data: You should compare the methods you learned from the research papers to the ones learned from lectures by implementing all methods to a specific data set and evaluating and comparing their performance by some clearly defined evaluation metrics. You can use one of the data sets posted on LEARN or any public data sets (for example, the ones on UCI Machine learning Repository or in R packages) subject to the instructor’s approval. In either case, a clear description of the data set is expected.
The main paper we worked with
- Bertsimas D., Gibson E., Dunn J., and Orfanoudaki A. 2022. “Optimal Survival Trees.” Machine Learning 111: 2951–3023.