Data mining air quality data for Athens, Greece
Quantitative data-driven decision support models are challenged by the difficulties in handling dynamic and uncertain features of real-world environmental systems. In addition, conditions for environmental management keep changing with time, demanding periodically updated decision support. These properties can be realized by learning from data, using knowledge discovery techniques. In the present paper, data mining techniques are applied for data analysis and for the construction of forecasting modules towards decision making, on the basis of air quality information for Athens, Greece. A number of data mining algorithms have been applied for the construction of forecasting models concerning maximum per day hourly ozone concentration values, for a total of 15 monitoring sites in Athens, Greece. In order to perform the experiments concerning the forecasting capabilities of the selected algorithms, two sets of ozone limit values were applied: the one resulting from the EU experience and practice, following the relevant legislation, and the other resulting from the detailed analysis and classification of the data. Successful forecasts are up to 95 demonstrating a good performance that should be considered for air quality forecasting modules applied at an operational basis.
M. Efraimidou, M. Kanaki, I. N. Athanasiadis, P. Mitkas, K. Karatzas, Data mining air quality data for Athens, Greece, Managing Environmental Knowledge (Proc. 20th Int'l Conference on Informatics for Environmental Protection: EnviroInfo 2006), pg. 505-508, 2006, Shaker Verlag.
You might also enjoy (View all publications)
- Introducing digital twins to agriculture
- Location-specific vs location-agnostic machine learning metamodels for predicting pasture nitrogen response rate
- Machine learning for large-scale crop yield forecasting