The is starting a new open workgroup to collectively read, discuss and better understand machine learning methods. The rise of new forms of data and the democratisation of computing power have reshaped the quantitative landscape in a very short span of time. In order to leverage their full potential, researchers need to tap into new methods that are not part of the usual quantitative curriculae. The open workgroup will fill this gap by providing a space for interested students and researchers to come together, read, and discuss some of the statistical and computational methods that gained prominence in the last few years.
The workgroup is structured in monthly sessions which will revolve around a coherent topic or family of methods. Prior to each meeting, a list of readings will be proposed as a prerequisite for attendance. On the session, a PhD student working on the area will provide a 20/30 minute introduction which will lead into the discussion, moderated by the student and members of staff. Attendance is free and open to anyone, and the topics are independent, so one can skip a month and join afterwards. When attending, however, it is expected everyone will have read the materials and will be ready to participate as it is the only way to make the sessions lively and useful.
For the Autumn semester of 2017/18, the schedule is the following:
- , by Dani Arribas-Bel and Mark Green.
- Nov. 17th: Regression trees and boosting, by Alec Davies.
- Dec. 15th: “word2vec” and text analysis, by Sam Comber.
The group will meet at 3pm at the conference room in the fourth floor of the . Attendance is free but you will need to RSVP. Stay tuned for updates on each of the sessions (readings, etc.), which will be advertised on this blog.