Logo Utrecht University

Human Data Science (HDS)

Colloquium

19/10/2017 – Learning Analytics

Matthieu Brinkhuis gave an exciting lecture on data analysis of the dynamic maths learning platform maths garden. A great discussion ensued about the details of the models used and their connection to the real world: Announcement: In this MSDSlab meeting, we’ll have a guest lecture by our member Matthieu Brinkhuis,  Assistant Professor at Computer Science /…

Read more

05/10/2017 – Mutual Information

After discussing the source and uses of Mutual Information, Erik-Jan introduced the concept of Maximal Information Criterion (MIC): a relatively new measure of dependence designed to detect both linear and nonlinear dependencies between two variables. We held a competition to find a simple (maximum 2 operations) functional relationship that the Maximal Information Coefficient could not…

Read more

Summer Break

MSDSlab was on summer break and returned at the start of the 2017/2018 academic year.

Read more

08/06/2017 – High-Performance Cluster Computing

Thursday 08/06/2017 at 14:00 in room B1.09 Oğuzhan Öğreden (MS Master) and Erik-Jan will be giving an explanation and demonstration of the UMC Utrecht High-Performance Cluster. It will be a practical session, where we will go over how to actually create, split, and submit simulations or heavy computing tasks to such a cluster. how create,…

Read more

11/05/2017 – Bayesian networks and sensitivity analysis

Thursday 11/05/2017 at 14:00 in room B1.09 In this MSDSlab meeting, dr. Silja Renooij (UU Computer Science, Decision Support Systems) will tell us about her research in Bayesian networks: Bayesian networks are flexible probabilistic models that can be used for reasoning under uncertainty. What can a sensitivity analysis of such a network tell us? See you…

Read more

13/04/2017 – Classifying Dutch Police Records

Thursday 13/04/2017 at 10:00 in room B1.09 In this MSDSlab meeting, Minke Brandenburg (UU, Artificial Intelligence) will tell us about her project on the topic of text mining in police records: Text mining can be very useful to perform text classification tasks. What is needed to build a model to classify Dutch police records? See…

Read more

30/03/2017 – The Intuition behind PageRank

Thursday 30/03/2017 at 14:00 in room B1.09 For this MSDSlab, Johnny van Doorn (PhD student at UvA) joined us to give a talk about PageRank. This comes from a talk he gave for a similar group in Amsterdam. Johnny discussed the main idea behind Google’s groundbreaking algorithm and go into some detail on how it can…

Read more

01/03/2017 Extending SEM for High-Dimensional data

Thursday March 2 at 15:00 – 16:00 in room A3.17   Erik-Jan and Daniel are working on an NWO Talent Grant full proposal, to be handed in by the 6th of March. Part shameless self-interest, part compelling discussion, this meeting will contain an overview of this project, which falls squarely in the MSDSlab topic area. With ample room for discussion,…

Read more

02/02/2017 – Computational social science

Thursday 2 February, 11:00 room A3.17 At our next meeting Joop Hox will discuss “computational social science” based on an article written for the journal Methodology. Joop’s article (draft, not for circulation) can be downloaded here: hox_cssm_printversie. We will also evaluate how things are going with MSDSlab and discuss possible changes to or extensions of our…

Read more

15 December 2016: Comparing predictive model performance on reconviction data

Nikolaj discussed his 2014 paper “Which method predicts recidivism best?: a comparison of statistical, machine learning and data mining predictive models”, published in the Journal of the Royal Statistical Association, Series A. For those who missed Nikolaj’s interesting talk, he presented a similar talk for the RSS journal club, shown below. The meeting discussed the…

Read more