Table 3 Model adaptation results for predicting TV concentrations from the three treatment processes (influent, aerobic, sand) and across different WWTPs (Plant A, Plant H, Plant P1, Plant P2)
From: Viral particle prediction in wastewater treatment plants using nonlinear lifelong learning models
Model | Set | Influent | Aerobic | Sand | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
R² | RMSE | MAE | MSE | R² | RMSE | MAE | MSE | R² | RMSE | MAE | MSE | ||
LSTM | Train | 0.903 | 0.118 | 0.074 | 0.014 | 0.903 | 0.074 | 0.049 | 0.075 | 0.912 | 0.107 | 0.075 | 0.012 |
Test | 0.861 | 0.136 | 0.081 | 0.019 | 0.835 | 0.143 | 0.092 | 0.020 | 0.874 | 0.061 | 0.044 | 0.004 | |
Adaptation 1 (A) | 0.911 | 0.109 | 0.077 | 0.013 | 0.882 | 0.131 | 0.094 | 0.018 | 0.903 | 0.106 | 0.077 | 0.013 | |
Adaptation 2 (H) | 0.956 | 0.036 | 0.027 | 0.002 | 0.840 | 0.119 | 0.091 | 0.015 | 0.956 | 0.021 | 0.014 | 0.001 | |
Adaptation 3 (P1) | 0.944 | 0.071 | 0.054 | 0.006 | 0.902 | 0.059 | 0.041 | 0.004 | 0.886 | 0.094 | 0.070 | 0.009 | |
Adaptation 4 (P2) | 0.852 | 0.038 | 0.029 | 0.002 | 0.837 | 0.153 | 0.108 | 0.025 | 0.874 | 0.061 | 0.044 | 0.004 |