Search

It is challenging to use conventional one-variable-at-time (OVAT) batch experiments to evaluate multivariate/inter-parametric interactions between physico-chemical variables that contribute to the adsorptive removal of contaminants. Thus, chemometric prediction approaches for multivariate calibration and analysis reveal the impact of multi-parametric variation on the process of concern. Hence, we aim to develop an artificial neural network (ANN), and stepwise regression (SR) models for multivariate calibration and analysis utilizing OVAT data prepared through experimentation. After comparing the models' performance, ANN was the superior model for this application in our work. The standard deviations (SD) between the observed and ANN-predicted values were very close. The average correlation coefficient (R2) between observed and ANN-predicted values for the training dataset was 96.9%. This confirms the ability of our developed ANN model to forecast lindane removal accurately. The testing dataset correlation coefficients (89.9% for ANN and 67.75% for SR) demonstrated a better correlation between observed and predicted ANN values. The ANN model training and testing dataset RMSE values were 1.482 and 2.402, lower than the SR values of 4.035 and 3.890. The MAPE values for the ANN model's training and testing datasets, 0.018 and 0.031, were lower than those for the SR model. The training and testing datasets have low RSR and PBIAS values, implying model strength. The R2 and WIA values are above 0.90 for both datasets, proving the ANN model's accuracy. Applying our developed ANN model will reduce the cost of removing inorganic and organic impurities, including lindane, and optimize chemical utilization. [ABSTRACT FROM AUTHOR]

It is challenging to use conventional one-variable-at-time (OVAT) batch experiments to evaluate multivariate/inter-parametric interactions between physico-chemical variables that contribute to the adsorptive removal of contaminants. Thus, chemometric prediction approaches for multivariate calibration and analysis reveal the impact of multi-parametric variation on the process of concern. Hence, we aim to develop an artificial neural network (ANN), and stepwise regression (SR) models for multivariate calibration and analysis utilizing OVAT data prepared through experimentation. After comparing the models’ performance, ANN was the superior model for this application in our work. The standard deviations (SD) between the observed and ANN-predicted values were very close. The average correlation coefficient (R2) between observed and ANN-predicted values for the training dataset was 96.9%. This confirms the ability of our developed ANN model to forecast lindane removal accurately. The testing dataset correlation coefficients (89.9% for ANN and 67.75% for SR) demonstrated a better correlation between observed and predicted ANN values. The ANN model training and testing dataset RMSE values were 1.482 and 2.402, lower than the SR values of 4.035 and 3.890. The MAPE values for the ANN model’s training and testing datasets, 0.018 and 0.031, were lower than those for the SR model. The training and testing datasets have low RSR and PBIAS values, implying model strength. The R2 and WIA values are above 0.90 for both datasets, proving the ANN model’s accuracy. Applying our developed ANN model will reduce the cost of removing inorganic and organic impurities, including lindane, and optimize chemical utilization.

• Drought loss models of wheat are built using two different statistical models. • The important SPEI variables recognized by two methods are different. • The random forest-derived model outperforms the stepwise regression model. Agricultural drought loss models provide services for the rapid risk assessment of agricultural disasters, and regional disaster prevention and mitigation efforts. This paper takes wheat as an example, and chooses counties dominated by rain-fed farmland in Henan Province as the study area. Counties dominated by rain-fed farmland are determined by setting a rain-fed threshold that is related to the proportion of the effective irrigation area to the cultivated land area. Modeling samples are screened by considering both drought occurrence time and wheat yield reductions. Under different thresholds (30%, 40%, 50% and 60%), we use the yield loss ratio as the dependent variable and 24 standardized precipitation evapotranspiration index parameters as independent variables to build drought loss models using both a multivariate stepwise regression model and a random forest model. Yield loss ratio from 1990 to 2015 is calculated by decomposing historical wheat yield time series. 24 standardized precipitation evapotranspiration index variables are 1–3 months' time scale standardized precipitation evapotranspiration index during the growth period (from October to May of the following year) of winter wheat in Henan Province. The results show that the random forest-derived model outperforms the stepwise regression model in all tests. The accuracy of all the models increases with an increase of the proportion of the rain-fed threshold. When the rain-fed threshold is 60%, the R2 values of the random forest model and the multivariate stepwise regression equation are 0.720 and 0.523, respectively. The validation results show that the mean absolute error and the root mean square error of the multivariate stepwise regression are 1.38 times and 1.31 times larger than the mean absolute error and the root mean square error from the random forests model. Moreover, both models identify that standardized precipitation evapotranspiration indices in October (sowing/planting stage) and February (overwintering stage) are important variables. However, the multivariate stepwise regression model fails to recognize the importance of standardized precipitation evapotranspiration indices during April–May (filling stage). [ABSTRACT FROM AUTHOR]