A Strategy for Finding the Optimal Scale of Plant Core Collection Based on Monte Carlo Simulation

Table 2

The formulas distilled by Eureqa based on the simulation results of CR.

Size^a	Formula	Error^b	𝑅^2c	FN^d
34	𝑓(𝑥,𝑦)=81.00+0.11𝑦+2.63log(𝑥𝑦)−62.06𝑦+472.15𝑥+𝑥𝑦+2.63log(𝑥𝑦)	0.075	0.9935	(12)
28	𝑓(𝑥,𝑦)=78.65+0.09𝑦+3.01log(𝑥𝑦)−60.77𝑦+345.54𝑥+𝑦+𝑥𝑦	0.075	0.9934	(11)
26	𝑓(𝑥,𝑦)=77.94+0.09𝑦+3.10log(𝑥𝑦)−53.22𝑦+323.37𝑥+𝑦	0.075	0.9934	(10)
24	𝑓(𝑥,𝑦)=70.56+0.09𝑦+4.38log(𝑥+𝑥𝑦)−29.63𝑦+310.22𝑥+𝑥𝑦	0.076	0.9932	(9)
20	𝑓(𝑥,𝑦)=92.17+4.16log(𝑦)−291.401.34+𝑥+2.90log(𝑦)	0.078	0.9926	(8)
18	𝑓(𝑥,𝑦)=69.88+𝑦−61.33𝑥+4.56log(𝑥𝑦−8.75)	0.083	0.9917	(7)
15	𝑓(𝑥,𝑦)=65.90+5.12log(𝑥𝑦−5.07𝑦−4.24)	0.090	0.9896	(6)
13	𝑓(𝑥,𝑦)=64.26+5.39log(𝑥𝑦−5.02𝑦)	0.114	0.9757	(5)
12	𝑓(𝑥,𝑦)=85.72+5.40log(𝑦)−133.68𝑥	0.120	0.9746	(4)
11	𝑓(𝑥,𝑦)=70.37+0.40𝑥+5.46log(𝑥𝑦)	0.161	0.9619	(3)
9	𝑓(𝑥,𝑦)=59.55+5.95log(𝑥𝑦)	0.180	0.9571	(2)
7	𝑓(𝑥,𝑦)=84.24+0.03𝑥𝑦	0.389	0.7222	(1)

aThe complexity of the formula; bthe error of the fitted formula; ^c𝑅²: the coefficient of determination; ^dFN: formula number.