Figure 6 The architectural diagram of the ANN model for predicting pKa values
Figure 7 ANN performance improvement diagram
Figure 8 shows the plot of outputs with respect to the targets for the training, validation and testing processes. The line is an ideal trendline in which the ANN predicted values should equal the experimental values. Roverall was found to be 0.99424 (Figure 8), as a result, the ANN predicted values were in good agreement with the experimental data.