All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
The revised paper meets the requirements of the reviewers and is suitable for publication.
[# PeerJ Staff Note - this decision was reviewed and approved by Jyotismita Chaki, a PeerJ Computer Science Section Editor covering this Section #]
According to the author's response, the author solved related problems, revised the grammatical errors in the revised paper, and clearly expounded the contribution of this paper.
No comment.
No comment.
According to the author's response, the author solved related problems, revised the grammatical errors in the revised paper, and clearly expounded the contribution of this paper.
The article meet our standards. The article include sufficient introduction and background to demonstrate how the work fits into the broader field of knowledge. Most of the problems have been solved.
In last version the article is the lack of sufficient experimentation to demonstrate the validity and applicability of the three methods. The author repeated the training procedures 5-10 times and reported the resulting statistics and aggregated the overall results in a separate “Discussion” section to provide a
insights regarding issues facing all of the types of inverse problems in their experiments.
The author maintain that the relative merits and the relative applicability of the various methods is not particularly understood, and it is precisely this relative performance which is the purpose and contribution of this paper.
The decision was made with the input of the two reviewers. I hope you will take these concerns into consideration, and I look forward seeing the revised paper.
[# PeerJ Staff Note: Please ensure that all review comments are addressed in a rebuttal letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate. It is a common mistake to address reviewer questions in the rebuttal letter but not in the revised manuscript. If a reviewer raised a question then your readers will probably have the same question so you should ensure that the manuscript can stand alone without the rebuttal letter. Directions on how to prepare a rebuttal letter can be found at: https://peerj.com/benefits/academic-rebuttal-letters/ #]
I suggest you clarify the innovation of this paper in the introduction, that is, the contribution made by this article.
Secondly, this article contains a large number of formulas. I hope you can explain the meaning of each symbol in the formula in detail, so as not to cause trouble to readers.
In addition, the figures in this paper need to be explained in more detail or clearer images are selected, as shown in FIG. 11
Formal results should include clear definitions of all terms and theorems, and detailed proofs.
This article explores deep learning methods for inverse problems. In order to facilitate readers to implement the methods used in this article, I suggest that you need to clearly list the hyperparameters required for deep learning training in the article.
The data set used in this study are available publically which is completely limited and imbalance.
In this article, you use various quantitative indicators to measure the advantages and disadvantages of various methods. I hope you can describe in detail the principle of these indicators and their relationship with the quality of the model.
For the overall beauty of the article, I suggest you use a unified style for the drawings of this article.
The methods used in this article are the results of early work. I suggest you use the latest achievements to enrich the work of this article, such as the work of the last year.
No comment. See the comments in the above three parts for details
The main innovation of this manuscript is to compare the robustness of the three categories, and report a statistical analysis of their differences, which is not novel enough. The method of this paper is not innovative enough. In fact, most of the work is done by combining other people’s methods. Authors need to highlight their innovative contributions.
Although the article is clear on the whole, I still hope you pay attention to the clarity of the article. There are numerous places in the article that say the same thing over and over again. I hope you can check and delete the redundant parts.
Pay attention to consistency to avoid confusion. Another obvious problem with this paper is the lack of sufficient experimentation to demonstrate the validity and applicability of the three methods.
I recommend you to read this manuscript: " Quadratic Residual Networks: A New Class of Neural Networks for Solving Forward and Inverse Problems in Physics Involving PDEs"Perhaps the author can find inspiration by reading more literature in this field to further optimize this paper.
Please check the normalization and accuracy of the figure in the paper carefully. Many figures are far away from the analysis paragraphs. I hope they can be adjusted.
The literature on deep learning methods for solving inverse problems was classified into three categories, of which was evaluated on sample inverse problems of different types. Why not try more categories? You can make some experiments on more categories.
One obvious problem with this paper is the lack of enough experimentation to demonstrate the validity and applicability of the proposed method. The author needs to do more experiments with more angles and show them in this paper.
No comments
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.