Analyzing the Impact of Airborne Particulate Matter on Urban Contamination with the Help of Hybrid Neural Networks

Abstract

In this study, particulate matter (PM), total suspended particulate (TSP), PM 10 , and PM 2.5 fractions) concentrations were recorded in various cities from south of Romania to build the corresponding time series for various intervals. First, the time series of each pollutant were used as inputs in various configurations of feed-forward neural networks (FANN) to find the most suitable network architecture to the PM specificity. The outputs were evaluated using mean absolute error (MAE), mean absolute percentage error (MAPE), root mean square error (RMSE), and Pearson correlation coefficient (r) between observed series and output series. Second, each time series was decomposed using Daubechies wavelets of third order into its corresponding components. Each decomposed component of a PM time series was used as input in the optimal feed-forward neural networks (FANN) architecture established in the first step. The output of each component was re-included to form the modeled series of the original pollutant time series.The final step was the comparison of FANN outputs with wavelet-FANN results to retrieve the wavelet utilization outcomes. The last section of the study describes the ROkidAIR cyberinfrastructure that integrates a decision support system (DSS). The DSS system uses artificial intelligence techniques and hybrid algorithms for assessing children's exposure to the pollution with particulate matter, in order to elaborate PM forecasted values and early warnings.

Modelling Missing Data for PM2.5 Time Series Forecasting with Computational Intelligence

“…The development of a feed forward neural model for PM 10 forecasting in the city of Ploiesti (Romania) involved a data preprocessing phase in which the missing data were completed by interpolation of the existing measurements [6]. Also an interpolation method is used for filling the missing data, the obtained time series being the input for PM 2.5 concentration forecasting in Dalian, China with symbolic regression based on genetic programming [7].…”

Changes in Nutrient Concentrations in Shenzhen Bay Detected Using Landsat Imagery between 1988 and 2020

“…Similar to the neural network, the data were divided into a training set (80%) and test set (20%). To avoid overfitting, the data were subjected to a five-fold cross-validation [32]. The obtained C DIN and C PO4_P retrieval models yielded RMSEs of 0.571 and 0.032 and R 2 values of 0.66 and 0.80, respectively.…”