You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Referring to "Why the Fitness Function is not Called for Solution at Index 0?"
To force calling the fitness function for each solution in every generation, consider setting keep_elitism and keep_parents to 0. Moreover, keep the 2 parameters save_solutions and save_best_solutions to their default value False.
My project is to find best weights of neural net but subject to input change. So the fitness of the best elite is different in some generations. However, I want to have the feature of keep_elitism to maintain good candidate. How can I get with that?
The phrase Moreover, keep save_solutions = False and save_best_solutions = False. is ambiguous. I'm not sure that this is a must or a suggestion.?
I have created my own function to calculate entire population fitness is more efficient than calling fitness_func separately or in batch. Questions are.
When should I call my own function? (On_generation callback I guess)
Where should I store my returned fitness value? (I can only think of storing them in a global variable and retrieve from it when fitness_func is called
Keeping elitism means that latest fitness value of the elite will not be updated. How to solve this issue?