------------------------------------------------------------ Analysis of environmental data with machine learning methods May, 4 - 8. 2009 Tentative Course Schedule ------------------------------------------------------------ Monday, 4 MAY 2009 08:30 Registration 09:00 Introduction to knowledge discovery in environmental data Saso Dzeroski 10:30 Break 11:00 Introduction to machine learning Ivan Bratko 12:30 Lunch break 13:30 Induction of decision trees Ivan Bratko 14:15 Evaluating results of learning Ivan Bratko 15:00 Break 15:30 - 17:00 Hands-on exercises with WEKA (Demonstration of a classification tree induction program and practical work on environmental data) Tuesday, 5 MAY 2009 09:00 Induction of regression trees Rule Induction Bayesian classification (Naive Bayes) Saso Dzeroski 10:30 Break 11:00 Hands-on exercises with WEKA (Applying regression trees, rule induction, and Naive Bayes to environmental data) 12:30 Lunch break 13:30 Introduction to the nearest neighbor method Feature ranking, selection, and weighting Ensemble methods Saso Dzeroski 15:00 Break 15:30 - 17:00 Hands-on exercises with WEKA (Applying the nearest neighbor method, feature weighting/ranking/selection, and ensemble methods to environmental data) Wednesday, 6 MAY 2009 09:00 Machine learning and environmental/ecological modeling Saso Dzeroski 09:45 An introduction to equation discovery Sao Dzeroski 10:30 Break 11:00 Integrating background knowledge in equation discovery Ljupco Todorovski 12:30 Lunch break 13:30 Recent developments in equation discovery Ljupco Todorovski 14:15 Domain knowledge for modelling population dynamics of aquatic ecosystems Natasa Atanasova 15:00 Break 15:30 Case studies in automated modelling of population dynamics for aquatic ecosystems Natasa Atanasova 16:15 - 17:00 Habitat modeling for aquatic organisms Saso Dzeroski 19:00 Social dinner in downtown Ljubljana Thursday, 7 MAY 2009 09:00 Machine Learning Applications in Forestry (1) Forest border identification Brown bear habitat modeling Predicting forest properties using LANDSAT and LIDAR data Saso Dzeroski 10:30 Break 11:00 Machine Learning Applications in Forestry (2) Forest stand mapping from LiDAR data Habitat mapping from satellite images Climate change influence on habitats Andrej Kobler 12:30 Lunch break 13:30 Machine Learning Applications in Agriculture Influence of farming practices on soil fauna Predictive modeling for co-existence of GM and conventional crops (gene-flow, seedbank persistence, ferals) Marko Debeljak 15:00 Break 15:30 Participant presentations 17:00 Poster session (with drinks and snacks) Participants will have the opportunity to present their work briefly orally and in more detail through a poster Friday, 8 MAY 2009 09:00 Machine Learning Applications in Environmental Epidemiology Respiratory diseases in children Saso Dzeroski Habitat modeling for tick-borne disease vectors Marko Debeljak Effects of ocupational exposure to mercury Bernard Zenko 10:30 Break 11:00 Discussion (1) 11:45 Discussion (2) 12:30 Lunch 13:30 Discussion (3) 14:15 Machine learning application in disaster forecasting and relief Predicting runoff from a watershed Boris Kompare 15:00 Break 15:30 Predicting earthquakes Prediction of fire danger in the natual environment Saso Dzeroski 17:00 Closing remarks ------------------------------------------------------------