Big data in biomedicine. Big models?
Thursday 27th of February 2014, 9:45am - 6pm
Most statistical methods learn from data using probability models. Now that data has grown to become "big data" and the complexity of the inference is also bigger, should our models grow? If so, which of the old lessons still apply and which need to be revised?
The aim of the workshop is to provide a forum to discuss recent statistical modelling strategies to solve complex problems, with a focus on biomedicine, Bayesian methods and high-dimensional inference. Some specifics topics of interest are:
- Prior choice: objective approaches, model separation and the value of informative priors
- High-dimensional model selection, including graphical models and correlated data
- Bioinformatics and medical applications
- Guido Consonni (Universitá Cattolica di Milano). Presentation slides
- Mark Girolami (UCL and Warwick University). Presentation slides
- Luca La Rocca (Universita degli studi de Modena e Reggio Emilia). Presentation slides
- Jun Liu (Harvard University)
- Jianhua Hu (MD Anderson Cancer Center)
- Valen Johnson (Texas A&M University). Presentation slides
- Peter Müller (University of Texas, Austin). Presentation slides
- Donatello Telesca (University of California, Los Angeles)
- Rich Savage (Warwick University). Presentation slides
- Aki Vehtari (Aalto University). Presentation slides
Sessions 1-2 (9:30-11:30 and 11:50-13:00, respectively) will be in room LIB1. Room LIB1 is a lecture room on the ground floor of Library, Building No.34 on the Campus map, between the vending machines and library cafe.
Sessions 3-4 (14:00-15:50 and 15:50-16:00, respectively) will be in room L5. Room L5 is a lecture room on the main level floor of Science Concourse (between Chemistry and Engineering, Buildings No.9 and 16 respectively on the Campus map, across the bridge from Library)
Attendance is free, but you must register at the "Workshop Registration" tab. The capacity of the venue is limited, so we recommend you register as soon as possible.