The Fourth International Workshop on Environmental Applications of Machine Learning
September 27-29, 2004, Bled, Slovenia
http://www-ai.ijs.si/SasoDzeroski/ECEMEAML04/eaml.html
As environmental concerns grow and information technology develops, more and more data on the different aspects (physical, chemical, biological, ecological) of the environment are gathered. Important sources of such data include (but are not limited to) laboratory tests, environmental monitoring, and remote sensing. Important types of environmental data include temporal data and spatial data, which is often stored in geographical information systems (GIS).
There is an increasing need to analyze the collected environmental data for different purposes, which include the support for environmental management decisions. For example, the data can be used to build predictive ecological models for an ecosystem of interest. Also, patterns discovered in the data may help understand the environmental processes of interest by highlighting the interrelationships between different parameters. Machine learning and data mining can be used to discover patterns and build predictive models from environmental data.
EAML-04, The International Workshop on Environmental Applications of Machine Learning will provide a forum for presenting recent advances in applying machine learning and data mining techniques for the analysis of environmental data. Three workshops on the topic have been previously organized, attracting mainly researchers in ecological modelling interested in the use of machine learning. This workshop aims to bring together researchers from the areas of ecology, ecological modelling and environmental sciences, on one hand, and the areas of data analysis, data mining and machine learning, on the other. EAML-04 is organized jointly with ECEM-04, The Fourth European Conference on Ecological Modelling, during the week SEP 27 - OCT 1, 2004.
EAML-04 covers all topics related to the application of data mining and machine learning methods to environmental data. An indicative, but non-exhaustive list of topics is given below.
Analysis of environmental data with:
- computational scientific discovery
- decision and regression tress
- evolutionary computing (e.g., genetic algorithms and programming)
- statistical learning (e.g., kernel methods and SVMs)
- neural networks (e.g., MLPs and SOMs)
- probabilistic methods (e.g., Bayesian networks)
- relational learning methods
- rule induction methods
Data mining and machine learning for:
- modelling of different types of ecosystems:
- agricultural ecosystems
- aquatic ecosystems
- grassland ecosystems
- forest ecosystems
- modelling different aspects of ecosystems / ecological processes:
- biodiversity changes
- habitat suitability
- population dynamics
- analysis of different types of environmental data:
- monitoring data (air/soil/water samples; chemical, physical, biological)
- remote sensing data (e.g., on atmosphere, geology, vegetation)
- spatial data (e.g., GIS data on land cover)
- temporal data (e.g., time series data on pollution levels)
- various environmental applications of machine learning, e.g.,:
- for decision support in environmental management
- for earthquake prediction
- for environmental risk assessment
- in environmental epidemiology
- in meteorological/atmospheric sciences
- in predictive ecotoxicology
Abstracts (no longer than two pages) should be submitted by MAR 20, 2004. Notification of acceptance for presentation at the conference will be sent by APR 15, 2004. Full papers should be submitted by JUN 15, 2004. They will be reviewed by members of an international program committee for inclusion in a special issue of the journal Ecological Modelling and/or an edited book
|
