I'm using kfold for accuracy verification, but I have some questions
My code is as follows
kf = KFold(n_splits=5, shuffle=True, random_state=42)for train_id, test_id in kf.split(points):toolsbox.register("evaluate", model, points=train_points)
Each time I used toolsbox.register("evaluate", model, points=train_points) code to re-register his evaluation function, and I want to know if this works ?
Another question I have is whether individuals in the different folds mentioned above can cross-mutate, i.e., can individuals from two populations that differ only in the way the fitness function is computed cross-mutate in deap, and does the way the fitness function is computed have any effect on the cross-mutations of separate individuals?
I'm using kfold for accuracy verification, but I have some questions My code is as follows
kf = KFold(n_splits=5, shuffle=True, random_state=42)for train_id, test_id in kf.split(points):toolsbox.register("evaluate", model, points=train_points)Each time I usedtoolsbox.register("evaluate", model, points=train_points)code to re-register his evaluation function, and I want to know if this works ?Another question I have is whether individuals in the different folds mentioned above can cross-mutate, i.e., can individuals from two populations that differ only in the way the fitness function is computed cross-mutate in deap, and does the way the fitness function is computed have any effect on the cross-mutations of separate individuals?