Table 1 Comparison of MAPE value improvement rates among DNN, SVR and RF models under 12 characteristics
From: Prediction of fruit characteristics of grafted plants of Camellia oleifera by deep neural networks
model | Â | y1 | y2 | y3 | y4 | y5 | y6 | y7 | y8 | y9 | y10 |
|---|---|---|---|---|---|---|---|---|---|---|---|
RF | MAPE | -1.30 | -0.49 | 1.09 | 0.27 | -3.28 | 0.36 | 1.44 | 0.18 | 0.53 | -1.55 |
R2 | 0.13 | 0.19 | -0.27 | -0.25 | 0.21 | 0.00 | 0.00 | 0.05 | -0.13 | 0.33 | |
SVR | MAPE | -7.27 | -3.03 | 1.04 | 0.68 | 0.78 | -0.18 | 0.69 | -1.20 | 0.77 | 0.33 |
R2 | 0.13 | 0.19 | 0.17 | -0.33 | 0.01 | -0.10 | -0.22 | 0.05 | -0.28 | -0.09 |