MODULE 3 Data Mining and Knowledge Discovery – an introductory course with focus on veterinary epidemiology applications 4 day course June 23-26, 2013 Instructed by: Drs. Fernanda Dorea & Crawford Revie The increasing quantity of electronic data available (including clinical and laboratory data) is allowing epidemiological intelligence to move from hypothesis testing to knowledge discovery. Data mining technology allows researchers to analyse large observational data sets, previously collected for different purposes, to summarize the data in novel ways, and to find unanticipated relationships. This data-driven process adopts fundamentally different approaches to traditional question-driven methods. Data mining has been used in epidemiology for early detection of disease outbreaks (syndromic surveillance) and studies of antimicrobial resistance, among others. This four-day course provides an introduction to machine learning techniques, covering both unsupervised – those in which training datasets are not required – as well as supervised approaches to both classification and regression tasks in data mining. The main software used in the course will be RapidMiner, the most widely used open source package for data mining; some use of R will also be made to illustrate the integration of features not yet available within RapidMiner.