Colloquium
MSDSlab Meeting 06/10/2016
The first meeting we started with indicating our reasons for wanting to be involved in the MSDSlab. The ideas were diverse; a list of the different suggestions can be found in the comment to this post by Daniel.
We talked about the paper on Computational Social Science (http://science.sciencemag.org/content/323/5915/721.full), as well as the paper suggested by Emmeke on RE-EM trees for longitudinal and clustered data (http://link.springer.com/article/10.1007/s10994-011-5258-3). Because of the time constraint, we were not able to go in-depth on this topic. Hence, our next meeting will focus on this paper as well.
The introductory book by James, Witten, Hastie & Tibshirani was mentioned several times and can be found here: http://www-bcf.usc.edu/~gareth/ISL/index.html (free pdf download).
Before the meeting, Edith de Leeuw sent three articles she wrote on the topic of data science for Clou Today, a marketing research news outlet: 1, 2, 3 (Dutch)
See below the BAMI flowchart that Gerko showed during the meeting.
One Response to “MSDSlab Meeting 06/10/2016”
You must be logged in to post a comment.
We also discussed the interests of our members regarding data science, work they have done, and topics they would like to see addressed.
Topics mentioned:
– Methodological aspects of data science (several members)
– How can we interpret the results of “black box” analyses? What if the goal is not only prediction?
– What methods do we choose?
– Data mining/Machine learning for longitudinal data
– Big data as replacement or auxiliary for surveys
– Text mining
– “Computational social science” (not exactly DS, but overlapping)
– De-identification, privacy issues
– Methods from DNA expression analysis
– Network analysis
All members also expressed a keen interest in finding out what the commonly used methods and habits are in DS, and where the frontiers lie. Several people are interested in playing around with software to learn more about how it all works.
-Daniel