Marco Grzegorczyk (Science and Engineering)
Network reconstruction with Bayesian networks is a challenging and topical task in many scientific disciplines. In my Honour College course students without any background in Statistics learn to extract Bayesian networks from data. After the course many students would like to apply the gained expert knowledge to data from their subjects, but proper Bayesian network applications require sophisticated data-preparation steps. With the FIT grant I will employ a teaching assistant (TA) who will provide statistical consultancy. Together with the TA the students will perform case-studies whose goal is to analyse subject-specific data sets. The TA will assist the students performing the statistical analysis, writing the reports and preparing R software demos for illustration. We will strongly encourage the students to write the reports in terms of their own subject-specific vocabularies, so that the reports can be easily understood by their fellow students. The central questions will be: ‘How would you personally have liked to learn about this statistical problem? How do you think the problem and the solution can be best illustrated by an R demo?’. The major challenge will be to find the best trade-off between a non-mathematical (‘easy-to-understand’) language and mathematical correctness. To the best of my knowledge, there is no comparable course unit at RUG, where students without any statistical background can perform their own statistical case-studies.
|Last modified:||16 October 2017 11.24 a.m.|