IEEE Transactions on Neural Networks and Learning Systems

special issue, the entire submission process took nearly a year. The first review by three reviewers deemed the algorithm to be profound. After revisions, the second review resulted in rejection.

Upon receiving the second review, the AE did not provide specific comments from the reviewers. They only mentioned that one reviewer was satisfied with the review, another suggested improving the grammar, and the opinion of the third reviewer was not mentioned. I assume that there were no major issues, otherwise, it would have been mentioned. The main reason for rejection was that one AE believed that the modifications made during the second review did not meet their expectations. They mentioned some optimization issues with the algorithm but did not elaborate further, leaving one feeling speechless. This submission result inevitably leaves one feeling somewhat resentful.

This is another discouraging submission experience. I can understand the rejection of my previous submission on Auto, as there were some shortcomings. One reviewer thoroughly reviewed the entire article and provided many excellent suggestions, stating that these issues could be resolved. Although it was ultimately rejected, the reasons for rejection were valid and helpful for improving the manuscript. However, this time, I am at a loss as to how to make improvements.

I have also read many papers that did not pass the fellow review. There are some big shots in the group that I am familiar with, but this paper is one that I wrote in my spare time about MARL (Multi-Agent Reinforcement Learning), and everyone in the group is focused on communication and computer networks. Therefore, it did not pass the fellow review. Networking is one aspect, and on the other hand, I am also aware that the paper has some flaws. The review comments from a few peers were indeed very professional, accurately pointing out the flaws that I thought I had. As far as my own paper is concerned, I personally think it is similar to the review comments on ICLR in OpenReview. The average level of reviewers may not be as good as those at conferences because journals cover a wide range of content, and many people focus on AI applications. However, the high number of reviewers, especially in the popular direction of MARL in recent years, will always include very professional reviewers.