The use of decision trees in the modeling of a waste water treatment plant

Authors: Nataša ATANASOVA, Boris KOMPARE
Citation: Acta hydrotechnica, vol. 20, no. 33, pp. 351-370, 2002.
Abstract: Wastewater treatment plants (WWTP) are dynamic and complex systems, the management of which can be improved by different approaches to modeling and predicting their operation. Machine learning tools (decision trees) were used to build useful prediction models for wastewater treatment plant operation. The data base used for building the models is composed of measured quantitative as well as qualitative data on the WWTP. We were also provided with a microbiological analysis. The data are presented as a one-day situation of the plant operation. So far, classification of the data was made using the Linneo+ methodology. We extended the knowledge gained by classification by analyzing the classified data and constructing useful models that predict WWTP operation from inflow data. The WEKA program package, which includes most of the popular machine learning algorithms, was used for constructing the models.
Keywords: wastewater, wastewater treatment plant, modeling, machine learning, decision trees
Full text: a33na.pdf