ComBi
ComBi seminar on "Statistical aspects in epidemiology and health research"
Time: September 25th 14.00 - 16.30
Venue: University of Helsinki, "Chemicum", A.I. Virtasen aukio 1, Auditorium 110
This afternoon seminar is intended to bring together scientists working on problems in epidemiology and health research. The perspective of the talks is statistical, each presenting an application of state-of-the-art methodology in a real case study. Three invited senior scientists will give talks during the seminar. The speakers and their topics are:
Prof. Carlo Berzuini (Cambridge University):
Development and validation of a genome wide predictor of disease in coronary artery disease - a Sequential Monte Carlo approach
Home page: http://www.mrc-bsu.cam.ac.uk/BSUsite/AboutUs/People/carlo.xml
Dr. Sangita Kulathinal (University of Helsinki and Indic Society for Education and Development):
Case-cohort design in practice - experiences from the MORGAM Project
Abstract: When carefully planned and analysed, the case-cohort design is a powerful choice for follow-up studies with multiple event types of interest. While the literature is rich with analysis methods for case-cohort data, little is written about the designing of a case-cohort study. Our experiences in designing, coordinating and analysing the MORGAM case-cohort study are potentially useful for other studies with similar characteristics. The motivation for using the case-cohort design in the MORGAM genetic study is discussed and issues relevant to its planning and analysis are considered. Solutions are discussed for appending the earlier case-cohort selection after an extension of the follow-up period and for achieving maximum overlap between earlier designs and the case-cohort design.
Home page: http://www.inseed.org/h/b/kulathinals.html
Dr. Aki Vehtari (Helsinki University of Technology):
Gaussian process in health care data analysis
Abstract: The excellent databases currently available allow us to study several interesting aspects of the Finnish health care system. Some of these issues are inherently complex, and therefore correspondingly complex statistical models and computational methods need to be used for their analysis. Gaussian processes (GPs) can be used as a generic and flexible tool, to describe the prior knowledge that 'similar entities have generally similar properties'. For example, geographically nearby areas have generally similar mortalities, and individuals with similar histories of health care service usage tend to have similar risks of becoming institutionalized. In this talk, I will show how 'similarity' can be described with GPs, discuss the computational advantages and challenges, and illustrate these methods in the context of spatial epidemiology, personal risk assessment, length-of-stay prediction, and in clustering different types of health care users.
Home page: http://www.lce.hut.fi/~ave/
Huom. Tilaisuudessa on ensimmäisen luennon jälkeen lyhyt kahvitauko. Tarjoilua varten osanotosta pyydetään ilmoittamaan sähköpostilla Pekka Marttiselle osoitteella Pekka.Marttinen@helsinki.fi viimeistään tiistaina 23.9.