Tibor V Varga
Section of Epidemiology
Bartholinsgade 6, bg. 24, indgang Q, 1356 København K, CSS b. 24, Building: 24.2.xx
Primary fields of research
artificial intelligence, machine learning, predictive statistics, social inequalities, health justice, causal inference, biomarkers.
Fields of interest
- Artificial Intelligence, Machine learning
- Algorithmic Fairness
- Predictive statistics
- Social inequality, social inequity, social justice
- Causal inference
Tibor obtained his BSc in Biology (2009, ELTE, Hungary), MSc in Nutritional Sciences (2013, Semmelweis University, Hungary) and PhD in Genetic Epidemiology (2016, Lund University, Sweden) and is currently working as an Assistant Professor of AI and Causality at the University of Copenhagen.
During his PhD and previous postdoctoral fellowship, Tibor worked primarily with metabolic diseases, cardiovascular genetics, genetic association studies, gene x environment interactions and prospective omics analyses. He used a wide range of statistical and machine learning tecniques to gain information on associations and predictive biomarkers for various metabolic disease outcomes.
During his current Assistant Professorship, Tibor is working on how to use AI on prospective, life-course big data to fight social inequity and identify causal determinants of health and disease.
The last decade experienced a huge increase in the utilization of AI algorithms in healthcare and elsewhere. Initial high hopes for AI to reduce various human biases have come to a bit of a halt as it became apparent that AI often further propagates biases that are deeply embedded in the data it uses. While math does not care about sex, gender, skin color, sexual preference, religion and country of origin, our societies are far from equal and there is a high risk that any data collection might be sensitive to inherent societal biases and downstream analyses might further propagate these into results, solutions, policies, applications, publications, and in many ways normalize these inequalities. We need to understand this process, talk about this issue and fight inequality from the start. We need to assess our available data for biases, we need to use AI to remove biases and carefully assess algorithms and any results for fairness. The next couple of years will include a lot of exploration in these topics. Hope to make a difference, hope to educate and to generate a public conversation!