Table 6 Measured R2-scores of the target prediction models that were derived from three source feature extractors trained on the CMR-2D-PBG, CMR-2D-BG, and CMR-2D-GWBG datasets
Target Experimental Dataset | Best Competitor | NN-CroMEL | ||
|---|---|---|---|---|
fpbe | fhse | fgw | ||
ESTM-RT-PF | 0.688 (0.121) | 0.768 (0.116) | 0.775 (0.138) | 0.732 (0.099) |
ESTM-RT-ZT | 0.783 (0.080) | 0.856 (0.051) | 0.862 (0.043) | 0.833 (0.066) |
ESTM-HT-PF | 0.854 (0.097) | 0.945 (0.038) | 0.950 (0.044) | 0.934 (0.035) |
ESTM-HT-ZT | 0.787 (0.081) | 0.812 (0.117) | 0.827 (0.093) | 0.785 (0.144) |
SCMAT-BIN | 0.505 (0.063) | 0.608 (0.137) | 0.616 (0.098) | 0.602 (0.127) |
SCMAT-TER | 0.708 (0.075) | 0.826 (0.082) | 0.829 (0.045) | 0.823 (0.065) |
IPOP-200-QE | 0.685 (0.080) | 0.822 (0.065) | 0.824 (0.058) | 0.804 (0.066) |
IPOP-200-DT | 0.896 (0.108) | 0.908 (0.126) | 0.935 (0.093) | 0.930 (0.084) |
IPOP-300-QE | 0.630 (0.027) | 0.653 (0.115) | 0.670 (0.103) | 0.648 (0.122) |
IPOP-300-DT | 0.869 (0.041) | 0.833 (0.049) | 0.886 (0.058) | 0.841 (0.051) |
Liverpool-RT | 0.562 (0.108) | 0.656 (0.131) | 0.675 (0.138) | 0.674 (0.095) |
MGTT | 0.500 (0.209) | 0.703 (0.101) | 0.714 (0.099) | 0.725 (0.105) |
EFE | 0.883 (0.023) | 0.949 (0.033) | 0.952 (0.023) | 0.934 (0.047) |
EBG | 0.876 (0.018) | 0.951 (0.039) | 0.955 (0.028) | 0.958 (0.033) |
