Request for Updating MOEA/D-EGO Implementation in PlatEMO

[liazhao5]

It has come to our notice that the MOEA/D-EGO method proposed by Prof. Zhang in 2010, which is included in PlatEMO, has been frequently reported by users to contain bugs in its code. These issues often prevent users from reproducing the results as described in the original paper.

In light of these reports, we have taken the initiative to modify the existing code of MOEA/D-EGO as included in PlatEMO. Our revised version addresses the aforementioned issues, ensuring that users can achieve the expected results without encountering the previously reported problems.

We kindly request you to consider replacing the current implementation of MOEA/D-EGO in PlatEMO with our revised version. We have included our contact email in the new version so that we can continue to monitor and resolve any future issues reported by users.

In addition, we have a small suggestion: could we possibly add a directory .\PlatEMO\Algorithms\surrogate_models\ to store the codes for commonly-used surrogate models? PlatEMO already includes many model-based methods, and having different methods utilize the same codes for surrogate modeling would be more efficient and user-friendly.

[yhm-me]

I also support the suggestion and hope the author can provide a version that allows users to compare the differences between algorithms except for the surrogates.