24小时热门版块排行榜

返回列表

moriarty4869

新虫 (初入文坛)

应助: 0 (幼儿园)
金币: 0.5
帖子: 21
在线: 10.5小时
虫号: 2822053
注册: 2013-11-23
专业: 运筹与管理

[交流] 一审意见是大修，内附审稿意见，诸位大神帮忙看看发表的可能性大吗？已有4人参与

Reviewers' comments:

Reviewer #1: The paper is good and written well. The gaps in the original ××× are properly addressed. The following modifications are necessary.
1.    What is slice about the multidimensional objective function describing the trend of the function. This needs to be explained. What does this mean and what is the base of this statement.
2.    Page 3, paragraph 3: what is meant by 'minimizing every single agent'?
3.    The Rosenbrock function problems have also been solved in ×××× with different sampling spaces for individual agents. The performance comparison to these cases needs to be done as well.
4.    In detail discuss the difference between the neighborhood approach implemented in ×××  and the ×× method here. How they are different, advantages, limitations, etc. This is very important.
5.    I guess Askley function needs to be corrected to Ackley function.
6.    The use of dog-leg method needs to be justified.
7.    The possible disadvantages and limitations of the method in this paper must be listed. This will not reduce the novelty of the approach.

Reviewer #2: I am quite intrigued by this article. It appears to present an important advance to one type
of **, known as ***, by combining it with a sophisticated *** method.
However I cannot tell - the paper is too poorly written. So I cannot recommend that it be
published in its current form. Perhaps after a MAJOR revision.

1) First, the authors should have someone with better command of English proof-read the paper.

However there are lots of presentation problems that have nothing to do with command
of English:

2) As a minor matter, there are many weird formatting problems, like hyphenating words between
two consonants, having citations appear after a sentence, having figure captions too small
to read, etc. These just distract the reader, and are annoying.

3) Most importantly, much of the text is not understandable. E.g., the sentence starting
"Particularly, when gradients contain expectation or conditional expectation ...", the
paragraph starting "However, the adoption of Monte-Carlo sampling ..", the paragraph
starting "Note that the equation constraint ..", definition 2 (and more generally all
of the mathematics), the entire discussion on p. 6, etc.

4) There should be much more connection with the literature. For example, ** falls in
the same spirit as ***, ****, etc., and should be related to them. Similarly, it
not ** that ** should be compared to, but distributed ***. Another oversight is that the
authors only consider "****", without any discussion of the more
recent version of **, known as "****".

In summary, I think there is substantial advance buried within this paper that should
be made broadly known to the research community. However I cannot extract it by
reading the paper in its current form.

回复此楼