Validation of Large-Scale Classification Problem in Dendritic Neuron Model Using Particle Antagonism Mechanism
Round 1
Reviewer 1 Report
Three learning algorithms including back-propagation (BP), biogeography-based optimization, and competitive swarm optimizer (CSO) are used in the model of DNM to test the presented scheme. The similation results are fine. I recommend the acceptance of this paper after minor revision.
- Please check the format of the section and subsection, whether the number and symbol are correct or not such as 1, 1.1, A, (A).
- The Section Discussion is very short and the Section Results include many experimental results, the authors may consider how to balance those two sections. It is optional.
- One recent paper related to the optimization of large scale data, the author may cite this paper as following:
Shiwei Guan, Yuping Wang, and Haiyan Liu, “A New Cooperative Co-evolution Algorithm Based on Variable Grouping and Local Search for Large Scale Global Optimization,”Journal of Network Intelligence, Vol. 2, No. 4, pp. 339-350, Nov. 2017
Author Response
- Please check the format of the section and subsection, whether the number and symbol are correct or not such as 1, 1.1, A, (A).
Response: We have checked the format of sections and subsections of the manuscript to ensure they are correct.
- The Section Discussion is very short and the Section Results include many experimental results, the authors may consider how to balance those two sections. It is optional.
Response: As the content of “Discussion” is related to the appendix tables, it is recommended to be presented as a separate section in order to facilitate understanding and avoid confusion.
- One recent paper related to the optimization of large scale data, the author may cite this paper as following: Shiwei Guan, Yuping Wang, and Haiyan Liu, “A New Cooperative Co-evolution Algorithm Based on Variable Grouping and Local Search for Large Scale Global Optimization,”Journal of Network Intelligence, Vol. 2, No. 4, pp. 339-350, Nov. 2017.
Response: We have cited this paper as No. 36 in the “References” of manuscript:
Guan, S.W.; Wang, Y.P.; Liu, H.Y. A New Cooperative Co-evolution Algorithm Based on Variable Grouping and Local Search for Large Scale Global Optimization. Journal of Network Intelligence. 2017, 2, 339–350.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The manuscript, which has been submitted as a paper on Electronics Journal deals with the Artificial Intelligence.
Methods of artificial intelligence are very popular in different areas of our life. Especially the artificial intelligence is used in electronic, informatics, maths. In the paper the AI is used to process big data - classification and verification. The using of AI has influence on reducing of a cost of data processing.
Authors were carried out tests of three learning algorithms: DNM+BP, DNM+BBO, DNM+CSO.
In the paper three problems were considered: magic gamma telescope data set, stat log shuttle data set, skin segmentation data set. The problems are very specific. I think authors should analyse less complicated areas for comparison which also can include set of very big data. Can you write why such problems were considered?
Authors obtained very interesting results and they pointed that the DNM+BP is optimum in the case of execution time, DNM+CSO is the best to ensure both accuracy stability and execution time. DNM+BBO is a good solution taking into account the stability of comprehensive performance and convergence rate.
This application is now very important in many fields, and the subject is of high theoretical significance and has a large practical application value. Of course, it can also be of great interest to the readers of the MDPI Electronics Journal.
The work is clear and I judge it to be free from basic errors and faulty expressions. The given theory is well supported by many results including tests carried out.
For all the above, publication of this manuscript on the MDPI Journal Electronics is recommended.
Author Response
- In the paper three problems were considered: magic gamma telescope data set, stat log shuttle data set, skin segmentation data set. The problems are very specific. I think authors should analyse less complicated areas for comparison which also can include set of very big data. Can you write why such problems were considered?
Response: These three problems are the most downloaded open problems in the large-scale classification of UCI Machine Learning Repository, which are advanced and are in different fields. In order to increase the actual demand and application value of this article, we chose them as samples. And we have revised the following sentences of “Experiment” on pp. 6(L228-L229):
The most downloaded open data sets in different fields of UCI Machine Learning Repository are used [46]
Author Response File: Author Response.pdf
