Second meeting of the Groningen Institute of Applied Statistics (GIAS)

Background:

Groningen hosts a broad and rich expertise in applied statistics. However, this expertise is currently dispersed across different faculties and departments, which limits opportunities for collaboration, knowledge exchange, joint teaching, funding, and visibility. Unlike universities where applied statisticians are brought together in a central institute, Groningen lacks a common platform, despite having all the ingredients for one.

To help bridge this gap, statisticians from the faculties of Behavioural and Social Sciences, Science and Engineering, Medical Sciences, and Economics and Business launched the Groningen Institute of Applied Statistics (GIAS) last year. This institute is focused on using statistics and data science to solve real-world problems, and aims to connect and strengthen the applied statistics community in Groningen.

Programme:

• Talks: Spotlight on Research 

• Perspective: Teaching of Statistics and Data Science in Groningen

• Panel Discussion: A new Interfaculty Master’s Programme?
  

This second GIAS meeting will bring together statisticians and data-minded researchers from across the university. Two keynote speakers, Prof. dr. Gerard van den Berg (Economics and Business) and Prof. dr. Marco Grzegorczyk (Science and Engineering), will spotlight the research in statistics and data science within their respective faculties. Additional talks will address how statistics and data science are used and taught across the university. We will conclude with a discussion on the potential development of a new interfaculty Master's programme in Applied Statistics and Data Science.

More about the research done by the two keynote speakers:
Title: Sensitive ages for educational attainment and adult health: assessing the age-specific importance of economic and emotional conditions in childhood using a migrating-siblings design

Gerard van den Berg
Dept of Economics, University of Groningen
Dept of Epidemiology, University Medical Center Groningen 

Abstract:
Econometricians develop statistical methods for causal inference in conjunction with empirical analyses. This paper aims to be an example of this. Data on siblings who migrate as children in the same calendar year have been used to identify childhood "sensitive ages" at which conditions are strongly formative for adult outcomes. In reality, one may distinguish sensitive ages by the childhood exposure of interest, which could be economic conditions but also emotional or mental wellbeing, for example after the loss of a household member. We extend the "migrating siblings" design to incorporate this. In an empirical analysis we follow families migrating from deprived South Italy (Sicily, Sardinia, Naples) to rich Turin in the 1950s-1970s. Importantly, the father often migrated before the remainder of the family joined. This enables us to obtain insights on sensitive ages as characterized by economic conditions versus those characterized by emotional conditions. Adult outcome measures include completion of high school, hospitalization due to cardiovascular diseases and usage of antipsychotic drugs.

Title:
A New Bayesian Approach to Learning Hybrid Bayesian Networks}

Marco Grzegorczyk
Probability and Statistics group
Bernoulli Institute, FSE
University of Groningen

Abstract:
We propose a new approach for learning the structure of Bayesian networks from hybrid data, that is, data containing both continuous (Gaussian) and discrete (categorical) variables. Consistent with state-of-the-art hybrid Bayesian network models, we do not allow discrete variables to have continuous parent nodes. Our model differs from existing approaches by incorporating discrete variables through multivariate linear regression rather than mixture modeling. Specifically, we apply multivariate linear regression, using the discrete variables as potential covariates, to adjust the means of the continuous Gaussian variables while simultaneously learning the dependencies among them. For each continuous variable, we infer a separate regression model with its own set of covariates (discrete parent nodes).

At this interactive meeting, you will have the opportunity to meet people interested in statistics from different faculties, and to contribute to the future of statistics teaching at the RUG. Drinks and nibbles will be provided.

Everyone is welcome, but registration is required!