Academic writing (04)

1. However, these methods requires laborious and time-consuming experimental and analytical investigations.
2. The dimensionality of the original ORP dataset was \(600\times 102\) runs.
3. To identify the useful ORP data, the adjusted ORP data were separately inputted in different ranges.
4. To model the highly nonlinear heterotrophic microbial reactions in activated sludge, six machine learning models commonly used in nonlinear system.
5. Fig. 2 plots as the functions of RB-COD and SB-COD, The rate \(\mathrm{K_1}\) was strongly correlated with RB-COD, but weakly correlated with SB-COD.
6. Fig. 3 plots the ORP breakpoint time verses RB-COD and SB-COD.
7. Panels (c) and (d) plot the correlations versus the adjusted ORP decline rates.
8. That is, the ORP breakpoint time became the origin of the x-axis and Fig. 1 was redrawn as Fig. 2.
9. The ORP decline rates likely respond to the denitrification and fermentation rates, which themselves respond to the RB-COD and SB-COD, respectively.
10. The estimated RB-COD and SB-COD were largely contributed by the pre-breakpoint and post-breakpoint ORP data.
11. Fig. 4a plots the Pearson’s correlation coefficients between the RB-COD and SB-COd versus the adjusted ORP values in all runs. Fig 4.b plots the corresponding p values.