Next Article in Journal
The Supermoral Singularity—AI as a Fountain of Values
Next Article in Special Issue
The Emerging Role of Blockchain Technology Applications in Routine Disease Surveillance Systems to Strengthen Global Health Security
Previous Article in Journal
Multiparty Dynamics and Failure Modes for Machine Learning and Artificial Intelligence
Previous Article in Special Issue
Doppler Radar-Based Non-Contact Health Monitoring for Obstructive Sleep Apnea Diagnosis: A Comprehensive Review
Open AccessArticle
Peer-Review Record

Pruning Fuzzy Neural Network Applied to the Construction of Expert Systems to Aid in the Diagnosis of the Treatment of Cryotherapy and Immunotherapy

Big Data Cogn. Comput. 2019, 3(2), 22; https://doi.org/10.3390/bdcc3020022
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Big Data Cogn. Comput. 2019, 3(2), 22; https://doi.org/10.3390/bdcc3020022
Received: 6 March 2019 / Revised: 25 March 2019 / Accepted: 3 April 2019 / Published: 9 April 2019
(This article belongs to the Special Issue Health Assessment in the Big Data Era)

Round  1

Reviewer 1 Report

This is a third version of the manuscript. I was already pleased with the contents and presentation of the previous one. Some further improvements are shown here, and therefore I feel that the quality of the paper has been also enhanced.


Author Response

This is a third version of the manuscript. I was already pleased with the contents and presentation of the previous one. Some further improvements are shown here, and therefore I feel that the quality of the paper has been also enhanced.

Dear Reviewer.

Thank you very much for the comments.

We believe that the improvement of the paper is due to reviews as efficient as yours.

Thank you for the compliments.


Reviewer 2 Report

I have reviewed the manuscript "Pruning fuzzy neural network applied to the construction of expert systems to aid in the diagnosis of the treatment of cryotherapy and immunotherapy", Manuscript ID: BDCC-469134 that has been submitted for publication in the MDPI BDCC Journal and I have identified a series of aspects that in my opinion must be addressed in order to bring a benefit to the manuscript. In this paper, the authors develop a software tool designed to aid in the prediction of cryotherapy and immunotherapy treatments in the case of the Human papillomavirus (HPV) infection. I consider that the article will benefit if the authors take into account the following remarks and address within the manuscript the signaled issues:

1)      The sections of the manuscript. The manuscript under review will benefit if it is restructured in accordance with the MDPI BDCC Journal's Template that provides a more logical structure that is much more appropriate for a research article. The restructuring of the manuscript will also help the authors to better express the novelty of their work and the contribution that they have made to the current state of knowledge. Consequently, the manuscript under review should be restructured as follows: Abstract, Keywords, 1. Introduction, 2. Materials and Methods, 3. Results, 4. Discussion, 5. Conclusions (not mandatory), 6. Patents (not mandatory), Supplementary Materials (not mandatory), Author Contributions, Funding, Acknowledgments, Conflicts of Interest, Appendices and References.

2)      In the "Introduction" section, the authors did not identify a clear gap in the current state of knowledge that needs to be filled, a gap that is being addressed by their manuscript. The manuscript will benefit if the authors state clearly the gap in the current state of knowledge that their manuscript addresses. In addition to this, in the "Introduction" section the authors must also state directly the novel aspects of their work.

3)      The "Materials and Methods" section. It will benefit the paper if the authors restructure their paper and devise a proper "Materials and Methods" section, as requested by the MDPI BDCC Journal's Template. In order to bring a benefit to the manuscript, the authors should state and justify very clear in the "Materials and Methods" section, preferably within the first paragraph, the choices they have made when developing the final form of their proposed approach. In this section, the new developed methods should be described in detail while well-established methods (and information) can be briefly described and appropriately cited. In the "Materials and Methods" section, the authors should devise a flowchart that depicts the steps that they have processed in developing their research and most important of all, the final target.

4)      The equations within the manuscript. Just like when a portion of text that expresses an idea, a concept, a definition that has not been invented by an author has to cite and give appropriate credit to the original source, a mathematical formalism is subjected to the same requirements in a scientific article. If a definition, a statement is being expressed by using a mathematical formalism that was not brought to the current state of knowledge by the authors, the source must be cited. Consequently, as in the manuscript under review one can find equations that were not devised for the first time in the scientific literature by the authors, I consider that it is appropriately to cite their source (for example, equations (8)-(11)). In what concerns the equations that were devised by the authors for the first time, they should be explained in detail so that the reader can understand clearly how the authors arrived at the respective equations.

5)      In what concerns equation (1) on page 7, first of all, the uninorm definition depends on two variables denoted by x and y and therefore, in this equation the intervals for both variables must be specified in all the three cases. Secondly, for consistency reasons, in the case when x and y are between o and 1 one must correct the interval as being left-open and right-close (according to the cited paper [72]), otherwise the "o" value is contained by both intervals considered in the first two cases, namely [0,o] and [o,1]. Thirdly, in the last situation of the equation (1), the one when the value of the uninorm is expressed as "max(x,y) or min(x,y)", the authors must specify that this situation corresponds to the case "otherwise", not "otherside". Fourthly, I consider that the authors should change the notation for the identity element "o" that varies between 0 and 1 (maybe denoting it as "g" as in the cited paper [72]), as in the current form of the paper the notation "o" seems to be the zero figure.

6)      In what concerns equation (2) on page 7, the authors cite the paper (72) as being the source of this information, but the equation within the paper under review is different from the corresponding equation from the cited paper (as in the equation (2) from the manuscript under review w appears at both terms of the sum, while in the cited paper (72) w appears at the first term and w̅ (an overlined w) at the second one. Please revise this issue.

7)      Line 222: "…where z0 = 1, v0 is the bias, e zj e vj, j = 1, ...". What is the meaning of the letter "e" that is placed before zj and vj? Please revise this sentence.

8)      Page 8, after equation (5): "The columns of the matrix Z, defined in (5), correspond to the outputs of the hidden neurons of the SLFN with respect to the input…". Even if it is widely known, the SLFN acronym (as any other acronyms) should be explained the first time when it appears in the manuscript, for example under the form: "The columns of the matrix Z, defined in (5), correspond to the outputs of the hidden neurons of the Single Hidden Layer Feedforward Neural Network (SLFN) with respect to the input…".

9)      Page 11, Figures 6 and 7. In all the cases, the titles and the measurement units of both Ox and Oy axes are missing.

10)  The authors should pay more attention to the spelling, grammar and style as several errors have occurred. The authors must perform a thorough editing of their manuscript in order to achieve a proper readable paper. I am providing a few nonlimitative examples out of the many others that must be addresses within the manuscript:  on Page 4, at the caption of Figure 2: "…Avaliable in…" instead of "…Available in…"; at Line 307: "In this paper (FNN) thas use de unineuron and genfis1 [19]." - please revise this sentence, as in the current form its meaning is unclear.


Author Response

Dear Reviewer. Thanks for the contributions. The changes made are highlighted in bold below.


1)      The sections of the manuscript. The manuscript under review will benefit if it is restructured in accordance with the MDPI BDCC Journal's Template that provides a more logical structure that is much more appropriate for a research article. The restructuring of the manuscript will also help the authors to better express the novelty of their work and the contribution that they have made to the current state of knowledge. Consequently, the manuscript under review should be restructured as follows: Abstract, Keywords, 1. Introduction, 2. Materials and Methods, 3. Results, 4. Discussion, 5. Conclusions (not mandatory), 6. Patents (not mandatory), Supplementary Materials (not mandatory), Author Contributions, Funding, Acknowledgments, Conflicts of Interest, Appendices and References.


            Dear Reviewer. Thank you for the information.

             We follow the format of the latex emphasizing the main terms involved in each topic. This organization is widely used in articles of the fuzzy neural network area.

             We checked the last articles published in the journal, and we did not find the mentioned formats. Therefore, believing it is not a mandatory item, we maintain the current format. Thank you.

https://www.mdpi.com/2504-2289/3/1/19

https://www.mdpi.com/2504-2289/3/1/18


2)      In the "Introduction" section, the authors did not identify a clear gap in the current state of knowledge that needs to be filled, a gap that is being addressed by their manuscript. The manuscript will benefit if the authors state clearly the gap in the current state of knowledge that their manuscript addresses. In addition to this, in the "Introduction" section the authors must also state directly the novel aspects of their work.


Another reviewer also requested this information. We have made the necessary changes and additions to the introduction to highlight these items listed. Thank you very much.


3)      The "Materials and Methods" section. It will benefit the paper if the authors restructure their paper and devise a proper "Materials and Methods" section, as requested by the MDPI BDCC Journal's Template. In order to bring a benefit to the manuscript, the authors should state and justify very clear in the "Materials and Methods" section, preferably within the first paragraph, the choices they have made when developing the final form of their proposed approach. In this section, the new developed methods should be described in detail while well-established methods (and information) can be briefly described and appropriately cited. In the "Materials and Methods" section, the authors should devise a flowchart that depicts the steps that they have processed in developing their research and most important of all, the final target.


We appreciate the tips, but we believe that the organization made through the main concepts of the problem studied, the approaches carried out and the current works synthesize this information and make the paper more coherent with other works of the area. The text included some factors that were missing in the presentation of the methods, evidencing their differences.


4)      The equations within the manuscript. Just like when a portion of text that expresses an idea, a concept, a definition that has not been invented by an author has to cite and give appropriate credit to the original source, a mathematical formalism is subjected to the same requirements in a scientific article. If a definition, a statement is being expressed by using a mathematical formalism that was not brought to the current state of knowledge by the authors, the source must be cited. Consequently, as in the manuscript under review one can find equations that were not devised for the first time in the scientific literature by the authors, I consider that it is appropriately to cite their source (for example, equations (8)-(11)). In what concerns the equations that were devised by the authors for the first time, they should be explained in detail so that the reader can understand clearly how the authors arrived at the respective equations.


Thank you for your consideration. However, this model was elaborated in the paper called Pruning fuzzy neural networks based on unineuron for problems of classification of patterns. The equations, their formulations, were all debated in articles that worked on this line. In this paper, the focus was to apply the model with the variants to meet the specific problems. Equations 8, 9, 10 and 11 are commonly used to evaluate intelligent models. The quote was included.


5)      In what concerns equation (1) on page 7, first of all, the uninorm definition depends on two variables denoted by x and y and therefore, in this equation the intervals for both variables must be specified in all the three cases. Secondly, for consistency reasons, in the case when x and y are between o and 1 one must correct the interval as being left-open and right-close (according to the cited paper [72]), otherwise the "o" value is contained by both intervals considered in the first two cases, namely [0,o] and [o,1]. Thirdly, in the last situation of the equation (1), the one when the value of the uninorm is expressed as "max(x,y) or min(x,y)", the authors must specify that this situation corresponds to the case "otherwise", not "otherside". Fourthly, I consider that the authors should change the notation for the identity element "o" that varies between 0 and 1 (maybe denoting it as "g" as in the cited paper [72]), as in the current form of the paper the notation "o" seems to be the zero figure.


The changes were performed in the equation.


6)      In what concerns equation (2) on page 7, the authors cite the paper (72) as being the source of this information, but the equation within the paper under review is different from the corresponding equation from the cited paper (as in the equation (2) from the manuscript under review w appears at both terms of the sum, while in the cited paper (72) w appears at the first term and w̅ (an overlined w) at the second one. Please revise this issue.

The changes were performed in the equation.


7)      Line 222: "…where z0 = 1, v0 is the bias, e zj e vj, j = 1, ...". What is the meaning of the letter "e" that is placed before zj and vj? Please revise this sentence.


Typing error. It has been fixed.


8)      Page 8, after equation (5): "The columns of the matrix Z, defined in (5), correspond to the outputs of the hidden neurons of the SLFN with respect to the input…". Even if it is widely known, the SLFN acronym (as any other acronyms) should be explained the first time when it appears in the manuscript, for example under the form: "The columns of the matrix Z, defined in (5), correspond to the outputs of the hidden neurons of the Single Hidden Layer Feedforward Neural Network (SLFN) with respect to the input…".


Thank you. Adjusted.


9)      Page 11, Figures 6 and 7. In all the cases, the titles and the measurement units of both Ox and Oy axes are missing.


            Thanks for the tip.

             As they deal with histograms, where the y-axis is the frequency and the x-axis is the  term evaluated, we place the reference in the text.


10)  The authors should pay more attention to the spelling, grammar and style as several errors have occurred. The authors must perform a thorough editing of their manuscript in order to achieve a proper readable paper. I am providing a few nonlimitative examples out of the many others that must be addresses within the manuscript:  on Page 4, at the caption of Figure 2: "Avaliable in…" instead of "…Available in…"at Line 307: "In this paper (FNN) thas use de unineuron and genfis1 [19]." - please revise this sentence, as in the current form its meaning is unclear.


             Thanks for the contribution.

             Revisions have been made throughout the text.


Reviewer 3 Report

The article is at a decent level and topic is interesting. However, many aspects should be improved. So I can recommend to accept paper after major revision.

In order to improve manuscript, I suggest the following recommendations:

R1: Please write clearly in the abstract about: 1) background and objective (why this topic is important), 2) more details about the methods used, 3) results (present and summarize the obtained result), 4) conclusion.

R2: In section "Literature Review" please add that fuzzy systems can be effectively applied to problems from other fields, eg articles:

- "Approximation of phenol concentration using novel hybrid computational intelligence methods",

- "Classification of tea specimens using novel hybrid artificial intelligence methods",

- "Approximation of Phenol Concentration using Computational Intelligence Methods Based on Signals from the Metal Oxide Sensor Array",

- "Comparison of Artificial Intelligence Methods on the Example of Tea Classification Based on Signals from E-nose Sensors".

R3: In the end of introduction section, please write clearly: 1) what is the goal of the study, 2) list the innovative elements and 3) identify motivations for undertaking research.

R4: The description of the proposed method is unclear, please prepare a Figure (scheme) explaining the structure of FNN.

R5: Please delete not yours Figures 1,2,3.

R6: Why has not cross-validation been used? Please, think about repeating the calculations.

R7: Please add a description of what the colors mean, on Figures 6 and 7.

R8: Please, improve the quality of Figures 6,7,8,9.

R9: Please compare the proposed method with the methods of other authors (state of art comparison).

R10: Please correct English.

R11: Please write what is the difference between proposed method and FNN from Table 1 and 2.

R12: Please in the conclusion section: 1) list the advantages and disadvantages of the proposed solution, and 2) indicate the limitations of work.

R13: Please write what "Time" from Tables 1 and 2 mean? Is this the time of training or classifying?

R14: Please write how the structure of the proposed method was designed.

R15: Please provide values of all parameters of proposed method.

R16: Please specify how the parameters were selected.

R17: Please specify if the proposed method parameters were optimized. Please write how FNN parameters were optimized?


Author Response

Dear Reviewer.

Thanks for the contributions. Here are the comments on each one (bold text).


R1: Please write clearly in the abstract about: 1) background and objective (why this topic is important), 2) more details about the methods used, 3) results (present and summarize the obtained result), 4) conclusion.


Inclusions of the requested topics were included in the introduction to make reading easier.


R2: In section "Literature Review" please add that fuzzy systems can be effectively applied to problems from other fields, eg articles:

- "Approximation of phenol concentration using novel hybrid computational intelligence methods",

- "Classification of tea specimens using novel hybrid artificial intelligence methods",

- "Approximation of Phenol Concentration using Computational Intelligence Methods Based on Signals from the Metal Oxide Sensor Array",

- "Comparison of Artificial Intelligence Methods on the Example of Tea Classification Based on Signals from E-nose Sensors".


All articles were included in the literature review. Thanks for the valuable tips.


R3: In the end of introduction section, please write clearly: 1) what is the goal of the study, 2) list the innovative elements and 3) identify motivations for undertaking research.


Include: 

The objective of this study is to find a classifier to act in the identification of the feasibility of using cryotherapy or immunotherapy treatments according to the characteristics of patients with a high degree of accuracy.

The innovations proposed in this paper are in a hybrid architecture capable of combining efficient training of neural networks and fuzzy interpretation techniques to extract the most significant number of fuzzy rules from the database used in the problem.

It is expected that with the fuzzy rules extracted, the decision making will be more efficient and allow the treatments to be indicated to the correct patients.


R4: The description of the proposed method is unclear, please prepare a Figure (scheme) explaining the structure of FNN.


The authors preferred to explain the figure that deals with the hybrid network. The following text was included

where neurons A indicate fuzzy neurons generated by the ANFIS method, which are linked to Z neurons (unineuron) and which in turn are unified in a single neuron with a linear activation function. The first two layers of this model represent the fuzzy inference system. The third layer is the representation of a neural network of aggregation that has a single neuron.


R5: Please delete not yours Figures 1,2,3.


Dear Reviewer. Actually the images are not ours and so they came from a repository that allows the use of the images in scientific works.


R6: Why has not cross-validation been used? Please, think about repeating the calculations.


Dear reviewer

All tests were used with cross-validation procedures. For the definition of the parameters used k-fold as indicated in item 4.2. In this same item another cross-validation procedure is indicated, where we divide in each repetition 70% of the samples for training and 30% for test of the model. Samples were randomly selected in each test.


R7: Please add a description of what the colors mean, on Figures 6 and 7.


Dear reviewer

The description of the colors was already present in the text.

See: Figures \ ref {data} and \ ref {data}} respectively present the data collected by the researchers in immunotherapy and cryotherapy treatments, where red bars indicate success in treatment and blue bars that procedure was not efficient. The data were presented graphically using the weka software \ cite {hall2009weka}.


R8: Please, improve the quality of Figures 6,7,8,9.


Done.


R9: Please compare the proposed method with the methods of other authors (state of art comparison).


Dear Reviewer. I believe there is some mistake. The methods compared are from other authors and represent precisely the experiment proposed in the original article of the database, which is a state of the art for the problem. I did not understand the placement.


R10: Please correct English.


English was reviewed with the help of native speakers.


R11: Please write what is the difference between proposed method and FNN from Table 1 and 2.


Excuse me. We forgot to include the features of the model. Included in this release are:

The FNN model represents the architectural configuration used in \cite{guimaraes2018imuno}, using the same parameter values used in the proposed model, in addition to a decision consensus of 0.7 and the number of bootstrap replications in 16.


R12: Please in the conclusion section: 1) list the advantages and disadvantages of the proposed solution, and 2) indicate the limitations of work.


Included.

The advantage of the proposed method is the application of a hybrid technique capable of extracting knowledge from the database and improving the classification capacity of problems related to immunotherapy and cryotherapy.

As for disadvantages, it stands out the high time, compared to traditional neural networks models. The limitations of the model are related to the characteristics of the database, which should be exclusively numerical.


R13: Please write what "Time" from Tables 1 and 2 mean? Is this the time of training or classifying?


also, the time collected in the tables is given in seconds and involves the sum of the training time and testing of the models.


R14: Please write how the structure of the proposed method was designed.


The WEKA models were defined according to the original article, following the initial configurations of the tool.


R15: Please provide values of all parameters of proposed method.


As explained in the text, it originated from another paper.


R16: Please specify how the parameters were selected.


The parameters of the model proposed in the paper were defined by cross-validation


R17: Please specify if the proposed method parameters were optimized. Please write how FNN parameters were optimized?


Because it is an approach where the values of the neurons in the first layer are obtained by the ANFIS, the values of weights and bias of the unineurons are defined in a random way and the final calculation of the weights is done by ELM, it was not necessary to use optimization techniques .


Round  2

Reviewer 2 Report

I have reviewed the revised version of the manuscript "Pruning fuzzy neural network applied to the construction of expert systems to aid in the diagnosis of the treatment of cryotherapy and immunotherapy", Manuscript ID: BDCC-469134 that has been submitted for publication in the MDPI BDCC Journal and I can conclude that even if the authors did not address all the signaled issues, and have actually procrastinated, postponed some aspects, the manuscript has been improved.


Reviewer 3 Report

The authors have included all my comments so the article is ready to be published.

Back to TopTop