Universitair docent
Dr. M.R. (Marcel) Haas
Area(s) of expertise:
Data science in population health
Data science in population health
Introduction
I work in the trans-disciplinary team at the Health campus The Hague, where we investigate Population Health Management: a discipline where we study adverse health conditions in relation to the social, cultural and economic context. My role is the data (science) and the quantitative analysis. We use our ELAN data infrastructure where we combine data from many different sources (Statistics Netherlands for demography, general practitioners, hospitals, mental health institutions, Municap Health Services and local government for long term and youth care, and more), coupled at the individual level. I am active in the directions of Open Science and FAIR data.
I have a background in astrophysics and have ten years of experience doing data science in industry. I am also a beer sommelier.
I have a background in astrophysics and have ten years of experience doing data science in industry. I am also a beer sommelier.
Scientific research
Examples of research that I am involved in at the time of writing:
- Synthetic data generation as “Privacy Enhancing Technology”: we work with very sensitive data on the individual level, which results in serious practical burdens. Using realistic, but not real data accelerates research, enables students to use realistic data and helps making our science more open and our data more FAIR.
- Predictive modeling for the metabolic syndrome population, and for syndemic vulnerability and resilience at local community level.
- Meta-analyses of medical and educational research using advanced statistics and Large Language Models
- Data Science for Real World Evidence: how to deal with real world, i.e. “dirty” data (e.g. routine care data) in data science for population health.
- Synthetic data generation as “Privacy Enhancing Technology”: we work with very sensitive data on the individual level, which results in serious practical burdens. Using realistic, but not real data accelerates research, enables students to use realistic data and helps making our science more open and our data more FAIR.
- Predictive modeling for the metabolic syndrome population, and for syndemic vulnerability and resilience at local community level.
- Meta-analyses of medical and educational research using advanced statistics and Large Language Models
- Data Science for Real World Evidence: how to deal with real world, i.e. “dirty” data (e.g. routine care data) in data science for population health.
