Table 1 Comparison of multiple regression (REG), backpropagation neural network (BPNN), generalized regression neural network (GRNN) and support vector machine (SVM) based prediction accuracy of rice blast severity measured as correlation coefficient (r), coefficient of determination (r2) and percent mean absolute error (%MAE) of the observed value for 'cross-location' models over various years.
From: Machine learning techniques in disease forecasting: a case study on rice blast prediction
Year(s) | Location(s) | Multiple Regression (REG) | Artificial Neural Network (ANN) | Support Vector Machine (SVM) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Â | Â | Â | Â | Â | Â | BPNN | GRNN | Â | Â | Â | ||||
| Â | Training Data | Test Data | r | r 2 | %MAE | r | r 2 | %MAE | r | r 2 | %MAE | r | r 2 | %MAE |
2000 | L-I, L-II, L-III, L-IV | L-V | 0.62 | 0.38 | 75.12 | 0.51 | 0.26 | 51.94 | 0.62 | 0.38 | 37.83 | 0.62 | 0.38 | 37.42 |
| Â | L-I, L-II, L-III, L-V | L-IV | 0.59 | 0.35 | 99.77 | 0.69 | 0.48 | 72.75 | 0.60 | 0.36 | 41.39 | 0.69 | 0.48 | 39.41 |
| Â | L-I, L-II, L-V, L-IV | L-III | 0.57 | 0.33 | 39.20 | 0.60 | 0.36 | 49.47 | 0.75 | 0.56 | 40.35 | 0.84 | 0.71 | 23.71 |
| Â | L-I, L-V, L-III, L-IV | L-II | 0.63 | 0.40 | 62.92 | 0.78 | 0.61 | 26.60 | 0.89 | 0.79 | 25.81 | 0.94 | 0.88 | 17.29 |
| Â | L-V, L-II, L-III, L-IV | L-I | 0.44 | 0.19 | 58.01 | 0.48 | 0.23 | 57.62 | 0.51 | 0.26 | 55.62 | 0.54 | 0.29 | 50.88 |
2001 | L-I, L-II, L-III, L-IV | L-V | 0.39 | 0.15 | 89.27 | 0.66 | 0.44 | 59.17 | 0.95 | 0.90 | 95.08 | 0.98 | 0.96 | 66.40 |
| Â | L-I, L-II, L-III, L-V | L-IV | 0.78 | 0.61 | 59.77 | 0.79 | 0.62 | 57.34 | 0.89 | 0.79 | 33.11 | 0.95 | 0.90 | 24.16 |
| Â | L-I, L-II, L-V, L-IV | L-III | -0.24 | 0.06 | 73.18 | 0.21 | 0.04 | 59.45 | 0.41 | 0.17 | 50.50 | 0.51 | 0.26 | 35.30 |
| Â | L-I, L-V, L-III, L-IV | L-II | 0.50 | 0.25 | 52.35 | 0.45 | 0.20 | 59.50 | 0.52 | 0.27 | 48.87 | 0.63 | 0.40 | 42.56 |
| Â | L-V, L-II, L-III, L-IV | L-I | -0.27 | 0.07 | 115.20 | 0.38 | 0.14 | 96.66 | 0.34 | 0.12 | 94.98 | 0.41 | 0.17 | 76.22 |
2002 | L-I, L-II, L-IV | L-V | 0.51 | 0.26 | 94.75 | 0.86 | 0.74 | 84.29 | 0.93 | 0.87 | 82.64 | 0.82 | 0.67 | 36.09 |
| Â | L-I, L-II, L-V | L-IV | -0.71 | 0.50 | 110.72 | 0.72 | 0.52 | 79.11 | 0.72 | 0.52 | 57.87 | 0.99 | 0.98 | 51.54 |
| Â | L-I, L-V, L-IV | L-II | 0.61 | 0.37 | 76.15 | 0.64 | 0.41 | 69.48 | 0.69 | 0.48 | 67.63 | 0.81 | 0.66 | 45.36 |
| Â | L-V, L-II, L-IV | L-I | -0.15 | 0.02 | 108.31 | -0.34 | 0.12 | 143.29 | -0.30 | 0.09 | 127.32 | 0.12 | 0.01 | 95.39 |
2003 | L-I, L-II, L-III, L-IV | L-V | 0.46 | 0.21 | 67.92 | 0.50 | 0.25 | 53.61 | 0.63 | 0.40 | 49.30 | 0.86 | 0.74 | 42.93 |
| Â | L-I, L-II, L-III, L-V | L-IV | 0.14 | 0.02 | 74.96 | 0.46 | 0.21 | 65.17 | 0.51 | 0.26 | 59.29 | 0.86 | 0.74 | 41.80 |
| Â | L-I, L-II, L-V, L-IV | L-III | 0.53 | 0.28 | 68.09 | 0.59 | 0.35 | 58.59 | 0.87 | 0.76 | 49.57 | 0.88 | 0.77 | 19.76 |
| Â | L-I, L-V, L-III, L-IV | L-II | 0.69 | 0.48 | 69.78 | 0.70 | 0.49 | 61.70 | 0.72 | 0.52 | 58.39 | 0.78 | 0.61 | 53.17 |
| Â | L-V, L-II, L-III, L-IV | L-I | 0.57 | 0.33 | 53.05 | 0.65 | 0.42 | 51.09 | 0.69 | 0.48 | 49.85 | 0.70 | 0.49 | 49.18 |
2004 | L-I, L-II, L-IV | L-V | 0.18 | 0.03 | 90.55 | 0.63 | 0.40 | 61.81 | 0.73 | 0.53 | 41.01 | 0.83 | 0.69 | 38.86 |
| Â | L-I, L-II, L-V | L-IV | 0.68 | 0.46 | 78.63 | 0.72 | 0.52 | 45.20 | 0.76 | 0.58 | 37.10 | 0.86 | 0.74 | 31.67 |
| Â | L-I, L-V, L-IV | L-II | 0.39 | 0.15 | 89.98 | 0.10 | 0.01 | 65.18 | 0.44 | 0.19 | 51.82 | 0.69 | 0.48 | 46.54 |
| Â | L-V, L-II, L-IV | L-I | 0.47 | 0.22 | 54.88 | 0.43 | 0.19 | 69.34 | 0.67 | 0.45 | 67.79 | 0.74 | 0.55 | 40.70 |