Dear Arnab Majumder ,

I recommend to read the attached paper I think is largely sufficient in reply of you Question!

Best regards

Souhil

Well , thank you dear Professor Mouassa. I gathered many vital informations from this paper. But could you send me some materials from which I will be able to understand which phase corresponds exploration and exploitation for any metaheuristic in generally?

Looking forward for the reply.

Dear Arnab Majumder,

I suggest you to see links and attached files on topic.

https://www.springer.com/gp/book/9783319133584

https://link.springer.com/book/10.1007/978-3-319-89309-9

https://asdfdsdwew.pp.ua/forum/?q=Proceedings+of+the+International+Conference+on+Soft+Computing+for+Problem+Solving%28+SocProS+2011+December+20-+22%2C+2011

Article Exploration vs. Exploitation in Differential Evolution

Best regards

Interesting, exploration means search for new things.. exploitation destroy the known things...🤗

It is not easy to identify a single parameter that directly controls exploration and exploitation behaviour. Rather, it is an emerging behaviour of all the parameters used. In case of the Differential Evolution algorithm, the first thing you may change is the mutation strategy applied. For example, you can use "DE/rand/" strategies to increase exploration or use "DE/best/" to increase exploitation. Besides, you can increase the population size for better exploration. Other parameters such as CR and F also affect this balance, however this effect will be dependent on other parameters you decided before.

In Differential Search Algorithm, the constant values given in Scale, p1 and p2 are determinants to balance global (exploration) and local (exploitation) search capability of the algorithm.

In Differential Evolution, F plays key role in balancing the search whereas CR controls convergence speed.