Email - Research and Publication - Unissula

(1)

(2)

Email:

(3)

Dr.Hj. Sri Arttini Dwi Prasetyowati, M.Si <[email protected]>

Your paper will be included in Vol. 14 No. 1, March 2022

8 messages

[email protected] <[email protected]> Fri, Apr 1, 2022 at 8:33 PM To: [email protected]

Dear Author,

Your paper will be included in Vol. 14 No. 1, March 2022. Please fill in originality declaration and copyright transfer form as attached for authors and send back to us as soon as possible and append short CV and photograph of all authors.

The accepted paper for publication cost 3.750.000 IDR Bank BNI 46

Kampus ITB

No. Account: 0495624718 (IDR) Account Name: IJEEI

Paper ID : D21-10354 Category : G. Informatics

Title : Dataset Feasibility Analysis Method based on Enhanced Adaptive LMS method with Min-max Normalization and Fuzzy Intuitive Set

Institution : Universitas Islam Sultan Agung, Semarang, Indonesia1, Universitas Islam Sultan Agung, Semarang, Indonesia2, Universitas Islam Sultan Agung, Semarang, Indonesia3, Dian Nuswantoro University, Semarang, Indonesia4, Institut Teknologi Sepuluh Nopember Surabaya, Indonesia5

Author : Dr. Sri arttini dwi Prasetyowati1, Munaf Ismail 2, Eka Nuryanto Budisusila3, De Rosal Ignatius Moses Setiadi4, Mauridhi Hery Purnomo5

Email : [email protected] Thank you for your kind cooperation.

With best regards, The Secretariat

International Journal on Electrical Engineering and Informatics Institut Teknologi Bandung

Ganesha 10, Bandung 40132 Indonesia Email: [email protected]

Website: www.ijeei.org ODCT & APC.doc 38K

Dr.Hj. Sri Arttini Dwi Prasetyowati, M.Si <[email protected]> Sun, Apr 3, 2022 at 3:44 PM To: [email protected]

Dear Editor IJEEI,

Thank you very much for the good news.

I will send the originality declaration and copyright transfer form as soon as possible.

With Best Regards, Arttini

On Fri, Apr 1, 2022 at 8:34 PM <[email protected]> wrote:

Dear Author,

Your paper will be included in Vol. 14 No. 1, March 2022. Please fill in originality declaration and copyright transfer form as attached for

authors and send back to us as soon as possible and append short CV and photograph of all authors.

(4)

Author : Dr. Sri arttini dwi Prasetyowati1, Munaf Ismail 2, Eka Nuryanto Budisusila3, De Rosal Ignatius Moses Setiadi4, Mauridhi Hery Purnomo5 Email : [email protected]

Thank you for your kind cooperation.

Website: www.ijeei.org

--

Dr. Arttini DP., M.Si Prodi Teknik Elektro Fakultas Teknologi Industri

Univ. Islam Sultan Agung Semarang

ODCT & APC.doc 38K

Dr.Hj. Sri Arttini Dwi Prasetyowati, M.Si <[email protected]> Sun, Apr 3, 2022 at 4:21 PM To: [email protected]

Dear Editor,

Here I send a receipt for the accepted paper for publication cost 3.750.000 IDR on the attachment file.

Thanks Best regards, Sri Arttini Dwi P

Dear Author,

Kampus ITB

(5)

Email: [email protected] Website: www.ijeei.org

--

ODCT & APC.doc 38K

Dr.Hj. Sri Arttini Dwi Prasetyowati, M.Si <[email protected]> Mon, Apr 4, 2022 at 5:28 AM To: [email protected]

Bcc: "Dr.Hj. Sri Arttini Dwi Prasetyowati, M.Si" <[email protected]>, Lilik Wijoroso <[email protected]>

Dear:

The Secretariat

We send:

1.Originality Declaration and Copyright Transfer form 2. Short CV and photographs of all authors.

3. proof of payment of 3.750.000 IDR for publication cost Hopefully all of the requested files are complete.

Thank you for your attention.

With best regards, S. Arttini DP

Dear Author,

Kampus ITB

International Journal on Electrical Engineering and Informatics

(6)

Institut Teknologi Bandung

--

3 attachments

WhatsApp Image 2022-04-03 at 16.12.44.jpeg 63K

ODCT & APC ARTTINI D21-10354.pdf 276K

SHORT CV AND PHOTOGRAPH OF ALL AUTHORS D21-10354.docx 539K

[email protected] <[email protected]> Mon, Apr 4, 2022 at 8:12 AM To: "Dr.Hj. Sri Arttini Dwi Prasetyowati, M.Si" <[email protected]>

Pada 2022-04-03 15:28, Dr.Hj. Sri Arttini Dwi Prasetyowati, M.Si menulis:

Dear:

The Secretariat

We send:

1.Originality Declaration and Copyright Transfer form 2. Short CV and photographs of all authors.

3. proof of payment of 3.750.000 IDR for publication cost Hopefully all of the requested files are complete.

Thank you for your attention.

With best regards, S. Arttini DP

Dear Author,

Your paper will be included in Vol. 14 No. 1, March 2022. Please fill in

originality declaration and copyright transfer form as attached for authors and send back to us as soon as possible and append short CV and

photograph of all authors.

Kampus ITB

No. Account: 0495624718 (IDR)

(7)

Account Name: IJEEI Paper ID : D21-10354 Category : G. Informatics

Title : Dataset Feasibility Analysis Method based on Enhanced Adaptive

LMS method with Min-max Normalization and Fuzzy Intuitive Set Institution : Universitas Islam Sultan Agung, Semarang, Indonesia1, Universitas Islam Sultan Agung, Semarang, Indonesia2, Universitas Islam

Sultan Agung, Semarang, Indonesia3, Dian Nuswantoro University, Semarang, Indonesia4, Institut Teknologi Sepuluh Nopember Surabaya, Indonesia5

Author : Dr. Sri arttini dwi Prasetyowati1, Munaf Ismail 2, Eka Nuryanto

Budisusila3, De Rosal Ignatius Moses Setiadi4, Mauridhi Hery Purnomo5

Email : [email protected] Thank you for your kind cooperation.

Website: www.ijeei.org [1]

--

Univ. Islam Sultan Agung Semarang Links:

---

[1] http://www.ijeei.org --

--- Dear AUthor,

Thank you for your information Thank you for your kind cooperation.

Munaf Ismail <[email protected]> Mon, Apr 4, 2022 at 10:06 AM

To: "Dr.Hj. Sri Arttini Dwi Prasetyowati, M.Si" <[email protected]>

Alhamdulillah.. 🙏

Pada tanggal Min, 3 Apr 2022 4.05 PM, Dr.Hj. Sri Arttini Dwi Prasetyowati, M.Si <[email protected]> menulis:

--- Forwarded message --- From: <[email protected]>

Date: Fri, Apr 1, 2022 at 8:34 PM

(8)

Subject: Your paper will be included in Vol. 14 No. 1, March 2022 To: <[email protected]>

Dear Author,

Kampus ITB

--

(9)

Review: The paper presents a study on dataset feasibility analysis. The topic discussed in the paper is quite interesting, not many researchers conducted similar studies. However, the condition also raises question about the motivation. How significant the with regards to the similar conclusion that could be derived using much simpler method (e.g: simple statistical analysis on the dataset)? 1. What is dataset feasibility analysis? There is no single explanation about what feasibility analysis definition on the document. What is feasible dataset? What is decent dataset? 2. The auhors mentioned the dataset used is sourced from KEEL, importance of the study and in a part of the document it was mentioned there are 44 datasets. I could not find any other detail characteristics of the dataset. Several information especially related to the class distribution is very important. 3. The authors should consider to add any baseline methods to perform the experiment and provide analysis/discussion based on the experimental result. 4.

How to evaluate the feasibility analysis result? 5. "The dataset would be feasible to use if it has an uncertainty value under 1.5 and information value closed to 0.5". Are uncertainty value and information value the new metrics on dataset analysis introduced by the authors? Whether those threshold mentioned based on the experimental result or any other analysis? The authors also need to improve the writing. There are several typos and incomplete sentence. Please show all revised parts in the revised paper in blue sentences. Please carefully address the reviewer comments and explain all revision you have done and answer all reviewer comments or question in separate file.

Question before Number One:

How significant the importance of the study with regards to the similar conclusion that could be derived using much simpler method (e.g: simple statistical analysis on the dataset)?

Answer:

We highly appreciate for your attention to the importance of the study. We will add an explanation about "the importance of the study with regards to the similar conclusion that could be derived using much simpler method".

Research on checking the quality of the dataset has never been found, especially in prediction. Most statistical analysis on datasets was about to analyze the data and determine the type of distribution as Gaussian, uniform, or another distributions [1][2]. The importance of this study is to minimize the use of not feasible data in decision making. If the only data is data that is not feasible, then some of the analysis can be used to consider preprocessing. We have added this description in the paper on page 1.

Question 1:

What is dataset feasibility analysis? There is no single explanation about what feasibility analysis definition on the document. What is feasible dataset? What is decent dataset?

(10)

Answer Question 1:

We express our gratitude for the reviewer’s comment, so that we can learn more deeply about the meaning of dataset feasible analysis.

Dataset feasibility analysis means analyze the feasibility of the dataset that will be used for prediction process, whether the dataset is feasible for processing or not. The analysis is carried out using Least Mean Square Predictive, where the data has been normalized before.

The results of the Least Mean Square Predictive are used as input for the fuzzy intuitive set, and then the uncertainty values and information values are analyzed.

A prediction is strongly influenced by the feasibility of the dataset. If the dataset is bad, the prediction will be bad. With the dataset feasibility test, at least it can be determined the next steps that must be taken by the decision maker. For example, if there is an unbalanced dataset, it must be balanced first, or in other words, do preprocessing first so that the dataset becomes better. We have added this description in the paper on page 1.

Feasible dataset is a dataset that is feasible to be processed, that is, the prediction results have a fairly small error and have the certain value of Uncertainty and Information.

What we mean by Decent dataset is a feasible dataset, so we have replaced the phrase decent dataset with a feasible dataset.

Question 2:

The authors mentioned the dataset used is sourced from KEEL, and in a part of the document it was mentioned there are 44 datasets. I could not find any other detail characteristics of the dataset. Several information especially related to the class distribution is very important Answer Question 2:

Thank you very much for correcting. I do apologize. I am wrong. The truth is 44 data not 44 datasets (Page. 11).

Question 3:

The authors should consider to add any baseline methods to perform the experiment and provide analysis/discussion based on the experimental result.

Answer Question 3:

(11)

Thank you for the suggestion. We have taken references from several journals about the methods used in determining the accuracy of the four datasets used. We choose the datasets:

Ecoli0_vs_1, Glass 0, Hamberman and Pima because these four datasets have varying degrees of accuracy. Although the four selected datasets are declared accurate, they have different average accuracy. This is stated in the paper on page 11 of Table 4 and Table 5 [3]. Table 4 discussed about AUC (Area Under Curve) results for imbalanced datasets and Table 5 discussed about results of the Wilcoxon rank-sum test for AUC.

Table 4. AUC (Area Under Curve) Results for Imbalanced Datasets

Dataset

DGC+ DGC ADAC2 NN CSVM

C 4.5 CS

C 4.5 C 4.5 C 4.5

Average

None None CS CS CS RUS SMT SMT-

TL

Ecoli0_vs_1 0,9799 0,9642 0,9692 0,9796 0,9671 0,9832 0,9796 0,9832 0,9761 0,9757889 Glass0 0,865 0,8553 0,8101 0,6792 0,5074 0,8212 0,8206 0,7754 0,8039 0,7709 Haberman 0,6213 0,5062 0,5604 0,6245 0,5382 0,5752 0,6423 0,6539 0,6203 0,5935889 Pima 0,7394 0,5274 0,7114 0,7175 0,7289 0,7125 0,7235 0,7134 0,6948 0,6965333

From Table 4, the Haberman dataset had the lowest average of AUC, so that Haberman had the lowest accuracy. These results were in accordance with the results of this paper that Haberman had the highest uncertainty value, so that the accuracy was low.

Table 5. Results of the Wilcoxon rank-sum test for AUC

Dataset

DGC+ DGC ADAC2 NN CSVM

C 4.5 CS

C 4.5 C 4.5 C 4.5

Average

None None CS CS CS RUS SMT SMT-

TL

Ecoli0_vs_1 0,9652 0,9436 0,9305 0,9598 0,9479 0,9695 0,9598 0,9695 0,9504 0,9757889 Glass0 0,7037 0,7248 0,5812 0,2941 0,0181 0,5942 0,5942 0,5113 0,5431 0,5071889 Haberman 0,1977 0,0182 0,0946 0,2399 0,0840 0,1110 0,2614 0,2627 0,1787 0,1609111 Pima 0,4504 0,0682 0,3878 0,3943 0,4551 0,3976 0,4226 0,4064 0,3486 0,370111

While Table 5, it could be read that the accuracy for the Haberman data was the lowest, while Ecoli0_vs_1 had the highest average accuracy. This confirmed that the method presented in this paper, method that combined between Min-max Normalization, LMS and Fuzzy Intuitive set was the valid method. This method was also easy to use and robust.

Question 4:

How to evaluate the feasibility analysis result?

(12)

As explained in Answer Question 3, after it was concluded which data were feasible and which data were not feasible, then to ensure this research was correct, the results were compared with checking the accuracy of AUC (Area Under Curve) results for imbalanced datasets. It turned out that the results are very suitable, that is Haberman had the lowest average of AUC, so that Haberman had the lowest accuracy. These results were in accordance with the results of this paper that Haberman had the highest uncertainty value, so that the accuracy was low.

In addition, there are also results obtained from “Results of the Wilcoxon rank-sum test for AUC”. The accuracy for the Haberman data was the lowest, while Ecoli0_vs_1 had the highest average accuracy. This confirmed that the method presented in this paper, method that combined between Min-max Normalization, LMS and Fuzzy Intuitive set was the valid method. This method was also explained in this paper on page 11.

Question 5:

"The dataset would be feasible to use if it has an uncertainty value under 1.5 and information value closed to 0.5". Are uncertainty value and information value the new metrics on dataset analysis introduced by the authors? Whether those threshold mentioned based on the experimental result or any other analysis?

We thank you for your attention to details in uncertainty value and information value.

Uncertainty value and information value are not the new metrics on dataset analysis. From the Journal entitled “The fuzzy intuitive sets in the decision-making” mentioned the range of Uncertainty value and information value.

Uncertainty tolerance value H(𝛼) was 0 ≤ H(𝛼) ≤ 1,5; where the original formula is 0 ≤ H(𝛼) ≤ 𝑙𝑜𝑔₂3 equation (21) on page 7, while the tolerance value for information I(F) was -1 ≤ I(F) ≤ 1. The value of I(F) was positive and negative depending on the range of 𝜇(𝛼) [3].

Suggestion 1:

The authors also need to improve the writing. There are several typos and incomplete sentence.

Response Suggestion 1:

(13)

According to the reviewer's suggestion, we have improved the writing and used proofreading services for correcting grammar, spelling, and punctuation errors.

Suggestion 2:

Please show all revised parts in the revised paper in blue sentences. Please carefully address the reviewer comments and explain all revision you have done and answer all reviewer comments or question in separate file.

Response Suggestion 2:

Thanks for the hint. I will do it accordingly to the reviewer's instructions.

(14)

References (Additional)

[1] Rahul Varma, “Statistical analysis on a dataset you don’t understand,” Towar. Data Sci., 2020, [Online]. Available: https://towardsdatascience.com/statistical-analysis-on- a-dataset-you-dont-understand-f382f43c8fa5.

[2] B. Timothy C. Heeren, PhD, Professor of Biostastics, Jacqueline N. Milton, PhD, Clinical Assistant Professor, “Basic Statistical Analysis Using the R Statistical Package,” Bost. Univ. Sch. Public Heal., [Online]. Available:

https://sphweb.bumc.bu.edu/otlt/MPH-Modules/BS/R/R-Manual/R- Manual_print.html.

[3] J. Kulicka, “The fuzzy intuitive sets in the decision-making,” no. May, 2015, doi:

10.13140/RG.2.1.3608.4646.