ID 2520: Causal Survival for Treatment Prognosis with Continuously Updated Data
Master’s Thesis
Accurate survival prediction is essential for optimizing treatment strategies and clinical decision-making in patients with chronic diseases. Traditional survival models, such as Cox proportional hazards rely on static baseline covariates and do not account for evolving patient conditions. Advances in causal inference and machine learning enable dynamic survival models that adapt to longitudinal data and provide updated risk assessments over time.
Requirements
- Proficiency in programming languages such as Python or R.
- Strong background knowledge of machine learning and causal inference.
- Familiarity with longitudinal data analysis and time-to-event models.
- Independent and structured work approach with strong written communication skills.
Tasks
- Conduct a comprehensive literature review on dynamic survival models, causal inference in longitudinal data, and current methodological developments.
- Implement and validate various modeling approaches for handling time-varying covariates.
- Apply causal methods to estimate treatment effects and counterfactual scenarios.
- Document the research process, results, and insights in a comprehensive Master’s thesis, adhering to the standard thesis writing guidelines.
Supervisors
Please send me an email (tatiana.merzhevich@fau.de) with your resume and transcript of records to apply for the thesis.
Please send me an email (tatiana.merzhevich@fau.de) with your resume and transcript of records to apply for the thesis.