Leveraging DNA-Based Computing to Improve the Performance of Artificial Neural Networks in Smart Manufacturing

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper presents the effectiveness of dna-based computing (DBC) as a complementary approach. DBC is an innovative machine learning approach that is rooted in the central rules of partial biology and processes genetic information from DNA/RNA to egg protein. In this article, two machine learning approaches are considered. In the first approach, the ANN is trained and tested using time series datasets driven by long and short windows, and features are extracted from the time domain. DBC's ability to address the limitations of ANNs, especially for short-window-driven datasets, highlights its potential as a practical machine learning solution. The article is logically clear and the discussion is sufficient, and it is suitable for publication in this journal.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Recommendation: Reject

General Assessment:
While the manuscript explores an interesting concept—integrating DNA-Based Computing (DBC) with Artificial Neural Networks (ANNs) for pattern recognition in smart manufacturing—the study suffers from several critical issues that limit its scientific rigor and practical relevance.

Major Comments

1. Use of Artificial Dataset
   The study relies entirely on artificially generated datasets, which significantly undermines its applicability and generalizability. The authors do not provide a clear rationale for why synthetic data was used, nor do they explain what the data represents in a real-world context.

Recommendation: A real-world dataset should be used within a meaningful application context. Without this, the study lacks practical relevance and should not be accepted in its current form.

1. Lack of Methodological Clarity
   Section 2 is labeled as a “Literature Review,” yet it does not explain the core methodologies used in the study—namely, ANN and DNA-Based Computing. This omission makes it difficult for readers to understand the foundation of the proposed approach.

Recommendation: The authors must streamline the manuscript and provide concise, clear explanations of both ANN and DBC methodologies. Additionally, the authors must explain in the supporting information how to do it in MATLAB.

1. Dataset Description is Inadequate
   The manuscript fails to clearly describe the nature of the dataset. It is unclear what the data represents, how it was generated, and how it relates to real manufacturing processes.

Recommendation: The authors must explicitly describe the dataset’s structure, origin, and relevance. Additionally, the dataset should be made available as supplementary material to allow reproducibility.

1. Missing Implementation Details
   The manuscript does not provide sufficient information on how the ANN and DBC models were implemented in MATLAB.

Recommendation: Include supporting information or an appendix detailing the implementation steps, including code snippets or pseudocode.

1. Unclear Figures and Terminology
• Figure 7: It is unclear whether the variables labeled as Fl are datasets or features.
• Figures 8 and 9: The confusion matrices are shown for both training and testing phases, which is unusual. Typically, only testing (or validation) performance is reported, especially when using a train/test split.

Recommendation: Clarify the meaning of all variables and figures. Explain the rationale for presenting training confusion matrices and ensure consistency with standard machine learning evaluation practices.

• Figures out of contest

Figures 11 and 12 looks like out of the manuscript contest.

 

Minor comments

1. Overly Dense and Repetitive Language:
• The manuscript often repeats concepts (e.g., window size trade-offs) across multiple sections.
• Sentences are long and packed with jargon, which can obscure the main point.

Suggestion: Simplify and streamline the writing. Use shorter sentences and reduce redundancy.

1. Unclear Explanation of DBC:
• The DBC methodology is described using biological metaphors (DNA → RNA → Protein), but the actual computational steps are vague.
• The transformation from time series to DNA sequences to protein arrays is not intuitive for readers unfamiliar with bioinformatics.

Suggestion: Include a concrete example with a small dataset showing how a time series is encoded into DNA, then RNA, then protein, and finally into numerical features.

1. Figures Are Referenced but Not Explained Well:
• Figures like 5a and 5b are central to understanding DBC, but the text does not walk the reader through them clearly.
• The “protein-verse” network diagrams are visually interesting but conceptually unclear.

Suggestion: Add more descriptive figure captions and explain the figures step-by-step in the text.

1. Terminology and Notation:
• The use of symbols like Xl, Xs, Yl, Ys, Fl, Fs, etc., is consistent but overwhelming without a summary table.
• The mathematical notation is sometimes introduced without context.

Suggestion: Include a glossary or table summarizing all symbols and their meanings.

1. Lack of Practical Implementation Details:
• The paper does not provide enough information for replication (e.g., how DNA rules are defined, what software was used for DBC).
• The ANN architecture is briefly described but lacks justification for design choices.

Suggestion: Add a supplementary section or appendix with implementation details or pseudocode.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have nicely conglomerated the power of nucleic acid-based computing and machine learning to make a workable platform. However, before publishing the article following minor points can be taken into care:

The literature review is quite rigorous. However, they are discrete. The author can stitch the fragmented parts together to make a nice introduction.

The sections start with "This section presents and discusses....". The author can make the sections continuous for the sake of the readers.

As the authors have provided a nice literature ensemble, we can expect a nice pictorial representation of the standard nucleic acid-based computing scheme. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have improved the manuscript and addressed all my comments. Therefore, I recommend the publication of the manuscript in its present form

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have worked rigorously on the revision. The manuscript can be accepted in its current form.