Table 1 Table showing the overall prediction performance in function of the number of weights in the models.
Model | Seed Pearson (std) | Leaf Pearson (std) | # parameters |
|---|---|---|---|
Galiana | 0.202 (0.079) | 0.196 (0.090) | 1443747 |
transf50epochs | 0.202 (0.085) | 0.217 (0.093) | 1,443,747 |
transf30epochs | 0.180 (0.070) | 0.225 (0.091) | 1,443,747 |
frozen200 | 0.095 (0.069) | 0.100 (0.055) | 17,836 |
frozen500 | 0.105 (0.065) | 0.138 (0.057) | 44,536 |
BMSNN100 | 0.173 (0.084) | 0.216 (0.083) | 163,996 |
BMSNN50 | 0.172 (0.083) | 0.182 (0.097) | 98,396 |
GPSNN | \(-0.012\) (0.075) | 0.021 (0.051) | 21,036 |