hi,
I found that when the model predicts synthetic functions like ZDT2, the res_x values ultimately passed to res_y = task.predict(res_x) are res_x.max() = 2.2941973970327956 and res_x.min() = -0.8462044987253801. This is clearly problematic, as ZDT2 explicitly requires $x$ to be within the range $[0, 1]$ (see https://pymoo.org/problems/multi/zdt.html); this leads to inaccurate $y$ predictions and, consequently, an incorrect HV value. (seed = 0)
hi,
I found that when the model predicts synthetic functions like ZDT2, the$x$ to be within the range $[0, 1]$ (see https://pymoo.org/problems/multi/zdt.html); this leads to inaccurate $y$ predictions and, consequently, an incorrect HV value. (seed = 0)
res_xvalues ultimately passed tores_y = task.predict(res_x)areres_x.max() = 2.2941973970327956andres_x.min() = -0.8462044987253801. This is clearly problematic, as ZDT2 explicitly requires