A linear and non-linear polynomial neural network modeling of dissolved oxygen content in surface water: Inter- and extrapolation performance with inputs' significance analysis.

Accurate prediction of water quality parameters (WQPs) is an important task in the management of water resources. Artificial neural networks (ANNs) are frequently applied for dissolved oxygen (DO) prediction, but often only their interpolation performance is checked. The aims of this research, beside interpolation, were the determination of extrapolation performance of ANN model, which was developed for the prediction of DO content in the Danube River, and the assessment of relationship between the significance of inputs and prediction error in the presence of values which were of out of the range of training. The applied ANN is a polynomial neural network (PNN) which performs embedded selection of most important inputs during learning, and provides a model in the form of linear and non-linear polynomial functions, which can then be used for a detailed analysis of the significance of inputs. Available dataset that contained 1912 monitoring records for 17 water quality parameters was split into a "regular" subset that contains normally distributed and low variability data, and an "extreme" subset that contains monitoring records with outlier values. The results revealed that the non-linear PNN model has good interpolation performance (R2=0.82), but it was not robust in extrapolation (R2=0.63). The analysis of extrapolation results has shown that the prediction errors are correlated with the significance of inputs. Namely, the out-of-training range values of the inputs with low importance do not affect significantly the PNN model performance, but their influence can be biased by the presence of multi-outlier monitoring records. Subsequently, linear PNN models were successfully applied to study the effect of water quality parameters on DO content. It was observed that DO level is mostly affected by temperature, pH, biological oxygen demand (BOD) and phosphorus concentration, while in extreme conditions the importance of alkalinity and bicarbonates rises over pH and BOD.

[1]  Hikmet Kerem Cigizoglu,et al.  Generalized regression neural network in modelling river sediment yield , 2006, Adv. Eng. Softw..

[2]  An assessment of trace element contamination in the freshwater sediments of Lake Iznik (NW Turkey) , 2016, Environmental Earth Sciences.

[3]  Soichi Nishiyama,et al.  A comparative study of artificial neural networks and neuro-fuzzy in continuous modeling of the daily and hourly behaviour of runoff , 2007 .

[4]  A. Ivakhnenko.  Heuristic self-organization in problems of engineering cybernetics , 1970 .

[5]  J. Adamowski,et al.  Application of wavelet-artificial intelligence hybrid models for water quality prediction: a case study in Aji-Chay River, Iran , 2016, Stochastic Environmental Research and Risk Assessment.

[6]  Li-Chiu Chang,et al.  Estimating spatio-temporal dynamics of stream total phosphate concentration by soft computing techniques. , 2016, The Science of the total environment.

[7]  E. Tziritis,et al.  Assessing the hydrogeochemistry and water quality of the Aji-Chay River, northwest of Iran , 2016, Environmental Earth Sciences.

[8]  Salim Heddam,et al.  Generalized regression neural network-based approach for modelling hourly dissolved oxygen concentration in the Upper Klamath River, Oregon, USA , 2014, Environmental technology.

[9]  Jan Adamowski,et al.  Multi-step water quality forecasting using a boosting ensemble multi-wavelet extreme learning machine model , 2018, Stochastic Environmental Research and Risk Assessment.

[10]  Sung-Kwun Oh,et al.  The design of self-organizing Polynomial Neural Networks , 2002, Inf. Sci..

[11]  Frank T.-C. Tsai,et al.  Supervised committee machine with artificial intelligence for prediction of fluoride concentration , 2013 .

[12]  Frank T.-C. Tsai,et al.  Prediction and structural uncertainty analyses of artificial neural networks using hierarchical Bayesian model averaging , 2015 .

[13]  S. J. Farlow.  The GMDH Algorithm of Ivakhnenko , 1981 .

[14]  Davor Antanasijević,et al.  Modelling of dissolved oxygen content using artificial neural networks: Danube River, North Serbia, case study , 2013, Environmental Science and Pollution Research.

[15]  Ton J. Cleophas,et al.  Artificial Intelligence, Multilayer Perceptron Modeling , 2013 .

[16]  M. Popovici.  Nutrient Management in the Danube River Basin , 2014 .

[17]  Joon Ha Kim,et al.  Development of early-warning protocol for predicting chlorophyll-a concentration using machine learning models in freshwater and estuarine reservoirs, Korea. , 2015, The Science of the total environment.

[18]  C. Flinders,et al.  Quantifying Variability in Four US Streams Using a Long-Term Data Set: Patterns in Water Quality Endpoints , 2016, Environmental Management.

[19]  Miklas Scholz,et al.  Comparison of Relationships Between pH, Dissolved Oxygen and Chlorophyll a for Aquaculture and Non-aquaculture Waters , 2011 .

[20]  Rahim Barzegar,et al.  Forecasting of groundwater level fluctuations using ensemble hybrid multi-wavelet neural network-based models. , 2017, The Science of the total environment.

[21]  A. Grossman,et al.  Responses to Macronutrient Deprivation , 2009 .

[22]  H. Altun,et al.  Treatment of multi-dimensional data to enhance neural network estimators in regression problems , 2006 .

[23]  Viktor Pocajt,et al.  A GMDH-type neural network with multi-filter feature selection for the prediction of transition temperatures of bent-core liquid crystals , 2016 .

[24]  Rahman Khatibi,et al.  Groundwater vulnerability indices conditioned by Supervised Intelligence Committee Machine (SICM). , 2017, The Science of the total environment.

[25]  Viktor Pocajt,et al.  Modelling of dissolved oxygen in the Danube River using artificial neural networks and Monte Carlo Simulation uncertainty analysis , 2014 .

[26]  Sajjad Ahmad,et al.  Suspended sediment load prediction of river systems: An artificial neural network approach , 2011 .

[27]  Ali Talebi,et al.  Development of a Hybrid Wavelet Packet- Group Method of Data Handling (WPGMDH) Model for Runoff Forecasting , 2016, Water Resources Management.

[28]  Ruey-Shiang Guh.  EFFECTS OF NON-NORMALITY ON ARTIFICIAL NEURAL NETWORK BASED CONTROL CHART PATTERN RECOGNIZER , 2002 .

[29]  P. A. Araoye,et al.  The seasonal variation of ph and dissolved oxygen (dO2) concentration in Asa lake Ilorin, Nigeria , 2009 .

[30]  Kwang-Tsao Shao,et al.  A data-mining framework for exploring the multi-relation between fish species and water quality through self-organizing map. , 2017, The Science of the total environment.

[31]  Usha A. Kumar,et al.  Comparison of neural networks and regression analysis: A new insight , 2005, Expert Syst. Appl..

[32]  Ozgur Kisi,et al.  Modeling of Dissolved Oxygen Concentration Using Different Neural Network Techniques in Foundation Creek, El Paso County, Colorado , 2012 .

[33]  L. Anastasakis,et al.  The Development of Self-Organization Techniques in Modelling: A Review of the Group Method of Data Handling (GMDH) , 2001 .

[34]  Ali Danandeh Mehr,et al.  A comparative analysis among computational intelligence techniques for dissolved oxygen prediction in Delaware River , 2017 .