ࡱ >
\ p None B a = c @ = Z ?N*8 X" 1 Arial1 Arial1 Arial1 Arial1 Arial1 Arial1 Arial General `
Table_2 ` Dependent variable Modeling strategy
Method Calibration set Prediction set$ a r2 Std. b RMSE Std. r2
m
r2 rACC1 (pIC50) Nonlinear Stepwise-MLR-BRANN 0.711 0.040 0.372 0.042 0.710 0.087 0.641 0.084 0.422 0.085 GA-MLR-BRANN 0.750 0.340 0.035 0.733 0.106 0.665 0.102 0.414 0.057
SPA-MLR-BRANN 0.773 0.028 0.302 0.044 0.751 0.088 0.680 0.095 0.387 0.067 BRGNN 0.815 0.037 0.270 0.026 0.808 0.075 0.732 0.080 0.336 0.054 Linear Stepwise-MLR 0.607 0.045 0.460 0.025 0.617 0.065 0.560 0.062 0.533 0.060 GA-MLR 0.633 0.038 0.439 0.643 0.082 0.582 0.081 0.501 0.055 SPA-MLR 0.670 0.021 0.410 0.031 0.083 0.644 0.092 0.437 0.046 hACC2 (pIC50) 0.774 0.024 0.034 0.730 0.738 0.058 0.432 0.017 0.785 0.417 0.027 0.771 0.063 0.741 0.064 0.428 0.792 0.019 0.395 0.726 0.804 0.419 0.018 0.837 0.388 0.791 0.036 0.884 0.407 0.013 0.684 0.015 0.431 0.022 0.440 0.688 0.412 0.627 0.737 0.425 0.030 0.734 0.020 0.405 0.667 0.775 0.418 rACC1(LE) 0.574 0.033 0.003 0.566 0.099 0.540 0.105 0.007 0.008 0.524 0.114 0.052 0.004 0.584 0.056 0.594 0.010 0.622 0.006 0.620 0.041 0.630 0.570 0.011 0.561 0.550 0.577 0.023 0.005 0.522 0.049 0.039 0.555 0.602
d MbP?_ % * + &P &F