Application of ANN and SVM for prediction nutrients in rivers.

This paper presents the results of predicting nutrients in rivers on national level by the use of two artificial intelligence methodologies. Artificial neural network (ANN) and support vector machine (SVM) were used to predict annual concentration of nitrate and phosphate in rivers of eleven European countries. For creation of an optimal model of prediction, 23 industrial, economical and agricultural parameters were used for the period from 2000 to 2011. The data from 2000 to 2010 was used for training, while the data for 2011 was used for model validation. Optimization of different parameters of ANN and SVM was conducted in order to obtain the model with the best performances. Results of created models were evaluated by using statistical performances indicator named coefficient of determination (R2). The obtained results showed that ANN has better results in predicting nitrate and phosphate compared to SVM models. These results suggest that ANN model is a promising tool for prediction of nutrients in rivers.

Estimation of NMVOC emissions using artificial neural networks and economical and sustainability indicators as inputs.

This paper describes the development of an artificial neural network (ANN) model based on economical and sustainability indicators for the prediction of annual non-methane volatile organic compounds (NMVOCs) emissions in China for the period 2005-2011 and its comparison with inventory emission factor models. The NMVOCs emissions in China were estimated using ANN model which was created using available data for nine European countries, which NMVOC emission per capita approximately correspond to the Chinese emissions, for the period 2004-2012. The forward input selection strategy was used to compare the significance of particular inputs for the prediction of NMVOC emissions in the nine selected EU countries and China. The final ANN model was trained using only five input variables, and it has demonstrated similar accuracy in predicting NMVOC emissions for the selected EU countries that were used for the development of the model and then for China for which the input dataset was previously unknown to the ANN model. The obtained mean absolute percentage error (MAPE) values were 8 % for EU countries and 5 % for China. Also, the temporal trend of NMVOC emissions predicted in this study is generally consistent with the trend obtained using inventory emission models. The proposed ANN approach can represent a viable alternative for the prediction of NMVOC emissions at the national level, in particular for developing countries which are usually lacking emission data.

Modeling of ammonia emission in the USA and EU countries using an artificial neural network approach.

Ammonia emissions at the national level are frequently estimated by applying the emission inventory approach, which includes the use of emission factors, which are difficult and expensive to determine. Emission factors are therefore the subject of estimation, and as such they contribute to inherent uncertainties in the estimation of ammonia emissions. This paper presents an alternative approach for the prediction of ammonia emissions at the national level based on artificial neural networks and broadly available sustainability and economical/agricultural indicators as model inputs. The Multilayer Perceptron (MLP) architecture was optimized using a trial-and-error procedure, including the number of hidden neurons, activation function, and a back-propagation algorithm. Principal component analysis (PCA) was applied to reduce mutual correlation between the inputs. The obtained results demonstrate that the MLP model created using the PCA transformed inputs (PCA-MLP) provides a more accurate prediction than the MLP model based on the original inputs. In the validation stage, the MLP and PCA-MLP models were tested for ammonia emission predictions for up to 2 years and compared with a principal component regression model. Among the three models, the PCA-MLP demonstrated the best performance, providing predictions for the USA and the majority of EU countries with a relative error of less than 20%.