Health data science

Biostatistics, causal inference, statistical and mathematical modeling, massive data analysis, research on medical and administrative databases, data warehouse research

Presentation of the research field

Biostatistics, a discipline at the heart of biomedical research, structures the methodology of biological, clinical or epidemiological studies. This discipline has progressively integrated methods derived from information technologies to respond to the methodological problems posed by increasingly rich and complex data. Data Science, represented by the profession of data scientist, thus integrates modern statistical tools with algorithms derived from machine learning and artificial intelligence methods and data mining.
An emerging theme of research in biostatistics concerns the evaluation and development of innovative mathematical approaches to identify the CAUSAL effect of an exposure on a health event (e.g., the use of a treatment on survival), using observational data and without resorting to the experimental approach of drawing lots. The increasing availability of health data (i.e. medico-administrative data, electronic medical records, large cohort studies) opens up this possibility which is at the heart of the analysis problems encountered by all IPLESP teams.