Approaching Optimal Nonlinear Dimensionality Reduction by a Spiking Neural Network
Round 1
Reviewer 1 Report
The manuscript proposes a method for reducing the nonlinear dimensionality of complex databases by means of a spiking neural network.
Results clearly support the proposed approach.
However, the overall quality of presentation should be improved. Some detailed comments follow:
- line 3: add a comma after hardware
- line 7: add a comma after science
- line 12: ?? I cannot understand the onset of the paper "Both, present and following.." "Both" cannot stand alone
- line 31: Artificial Bee Algorithm should be Artificial Bee Colony
- line 45: PCA: maybe for the first occurrence it is better to define it "the classic Principal Component Analysis (PCA) ..
- lines 69-70: after doing a .. processes: here either you cancel "a" or it should be "process"
- lines 73-75: understandable.
- line 135: but optimizing? why you say "but"? Odd syntactic construction
- line 153: "Also because" I suggest you to use "In addition, "
- As a general comment: where Table x or Figure X is the subject of the phrase you cannot put a comma after the number of the Figure/Table.
- line 182: to writing of numbers --> to the writing of numbers
- line 233: in both networks, namely with ..
- line 246 (an the same is repeated again): As was proceeded in --> As it was proceeded in
- line 254 and 264: Why are you assessing that MSE values below 0.09 and 0.07 are readily acceptable?
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This paper is worth for acceptance, novelty of the idea seems interesting and small changes need to be incorporated in order to enhance.
1) This paper deals with an exciting topic. The article has been read carefully, and some crucial issues have been highlighted in order to be considered by the author(s).
2) All the acronyms should be defined and explained first before using them such that they become evident for the readers.
3) Most of the typos and incorrect grammars have been corrected, but it is still necessary to subject the paper to proofreading.
4) The paper needs to be restructured in order to be precise.The Introduction and related work parts give valuable information for the readers as well as researchers. In addition recent papers should be added in the part of related work.
5) Representation of figures needs to be improved.
6) Grammatical errors should be validated.
7) It would be good if security domains [1-2], such as adversarial examples, would be reflected in future research or related work.
[1] Kwon, Hyun, Hyunsoo Yoon, and Daeseon Choi. "Restricted evasion attack: Generation of restricted-area adversarial example." IEEE Access 7 (2019): 60908-60919.
[2] Kwon, Hyun, Hyunsoo Yoon, and Ki-Woong Park. "Acoustic-decoy: Detection of adversarial examples through audio modification on speech recognition system." Neurocomputing417 (2020): 357-370.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The paper proposes the use of spiking neural network for optimum dimensionality reduction. The effectiveness of the proposed technique is verified with simulations based on 3 kinds of datasets. However, the paper needs to address the following comments for publication.
- Detailed description on the goal of the paper is missing in Abstract and Introduction.
- Related Works section should be added to help the readers understand the past work and compare with the proposed work.
- More detailed analytical description on the reason for the dimensionality reduction improvement of the proposed technique compared to the conventional technique based on figures 6 to 14 is needed in Section 4 Experimental Results.
- More insight based on the experimental results should be added in Section 4.4 Discussion.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
The author have failed to address the review comments.
Author Response
Comments about the new manuscript for Reviewer 3
Page 1.
In the abstract, some lines were changed and added.
Page 2.
The marked lines were added.
Page 7.
The information was reordered omitting the initial Figure 2.
Page 10.
Subsection 4.1. is now dedicated to introducing the use of MSE values.
Page 13.
A brief statement for presenting Figure 7 is given, which is based on subsection 4.1.
Page 14.
A brief statement for presenting Figure 10 is given, which is based on subsection 4.1.
Page 16.
A brief statement for presenting Figure 13 is given, which is based on subsection 4.1.
Pages 16 and 17.
Item No. 3 was wholly rewritten.
Please see the new PDF version of the manuscript.
Author Response File: Author Response.pdf
