Peer Review Process

Machine Learning and Convergence Science (MLCS) applies a rigorous double-anonymous peer-review process to maintain the scientific quality, relevance, and integrity of its published content. Each manuscript that proceeds to external review is evaluated by at least two independent reviewers with expertise relevant to the methodological and application domains of the study. During peer review, reviewers do not receive information about the authors’ identities, and authors do not receive information about the reviewers’ identities. The editorial team maintains the confidentiality of manuscripts, reviewer reports, and editorial correspondence throughout the process.

1. Initial Editorial Screening

After submission, the editorial office conducts an initial administrative and technical screening. The journal checks the manuscript for completeness, adherence to the author guidelines, appropriate anonymisation, similarity with previously published material, and compliance with relevant ethical requirements. The Editor-in-Chief or an assigned handling editor then evaluates the manuscript for its relevance to the Aims and Scope of MLCS, scholarly contribution, methodological adequacy, clarity, and overall suitability for external peer review. The editor may reject a manuscript without external review when it falls outside the journal's scope, lacks sufficient methodological or substantive contribution, contains fundamental methodological weaknesses, does not meet basic reporting standards, or raises significant ethical or research-integrity concerns.

MLCS does not consider the application of an established machine-learning method to a new dataset, by itself, sufficient for publication. The manuscript should demonstrate a meaningful methodological contribution, substantive domain insight, or a well-justified integration of both.

2. Reviewer Selection

Manuscripts that pass editorial screening are normally assigned to at least two independent external reviewers. The handling editor selects reviewers on the basis of their subject expertise, publication record, methodological competence, and ability to evaluate the application domain addressed by the manuscript. When appropriate, the editor may invite reviewers with complementary expertise so that both the computational methodology and the substantive domain can be evaluated adequately. Reviewers should not have relationships that could reasonably compromise the independence of their assessment. Potential conflicts may include recent collaboration with the authors, institutional affiliation with an author, supervisory relationships, financial interests, close personal relationships, or direct involvement in the research under review. Authors may suggest or identify potential reviewers when requested by the journal; however, such suggestions are advisory. The editor retains full responsibility for reviewer selection.

3. Double-Anonymous Review

MLCS uses double-anonymous peer review. Authors must therefore remove names, affiliations, acknowledgements, and other directly identifying information from the manuscript submitted for review. The editorial office may return manuscripts that have not been adequately anonymised before initiating peer review. Editors make reasonable efforts to preserve anonymity. However, MLCS cannot guarantee complete anonymity when an author's identity can be inferred from preprints, public datasets, conference presentations, previous publications, or other publicly available information.

4. Review Criteria

Reviewers evaluate manuscripts according to their scholarly merit and relevance to MLCS. The assessment considers:

• originality and significance of the contribution;
• alignment with the aims and scope of MLCS;
• methodological rigour and appropriateness of the study design;
• validity of the application-domain assumptions and conclusions;
• adequacy of datasets, baselines, validation procedures, and statistical analyses;
• reproducibility and transparency of methods, data, code, and experimental procedures where applicable;
• appropriate consideration of uncertainty, robustness, calibration, explainability, fairness, privacy, or deployment when relevant to the study;
• clarity and accuracy of the presentation;
• adequacy and relevance of the literature review and references; and
• compliance with research and publication ethics.

For convergence research, reviewers should consider not only whether the computational method performs adequately, but also whether the study addresses a meaningful domain problem and whether its evaluation supports the scientific, clinical, operational, or policy claims made by the authors.

5. Reviewer Responsibilities and Confidentiality

Reviewers provide an objective, constructive, and evidence-based assessment of the manuscript. They should distinguish between essential revisions and optional recommendations and should avoid comments based on personal preference rather than scholarly considerations. Reviewers must treat the manuscript and all associated materials as confidential. They may not distribute unpublished material, use information obtained through peer review for personal advantage, or disclose manuscript content before publication. Reviewers must disclose any actual or potential conflict of interest before accepting a review assignment. When a reviewer believes that a conflict could affect, or reasonably appear to affect, their impartiality, the reviewer should decline the invitation. Reviewers must not upload manuscripts, supplementary materials, or confidential editorial information to publicly accessible generative AI systems or other third-party services that may retain, reuse, or disclose the material. Further provisions are described in the [Generative AI Policy].

These principles are consistent with COPE guidance emphasizing confidentiality, objectivity, competing-interest disclosure, and responsible conduct by peer reviewers.

6. Editorial Decision

Reviewers provide recommendations to assist the editor; reviewers do not make the final publication decision. After considering the reviewer reports, the scientific merits of the manuscript, its compliance with journal policies, and its alignment with the journal's scope, the handling editor makes or recommends one of the following decisions:

• Accept: the manuscript is suitable for publication with no substantive changes required.
• Minor Revision: the manuscript requires limited revisions that do not substantially alter its methods, results, or conclusions.
• Major Revision: substantial changes, additional analysis, clarification, or methodological improvement are required before the manuscript can be reconsidered.
• Reject: the manuscript does not meet the journal's standards or would require changes beyond the scope of a reasonable revision.

Editorial decisions do not depend solely on the numerical balance of reviewer recommendations. The editor evaluates the substance of each review and remains responsible for the final decision.

7. Conflicting Reviewer Recommendations

When reviewer recommendations differ substantially, the handling editor evaluates the reasoning and evidence provided in each report rather than resolving the decision by majority vote. The editor may request clarification from a reviewer, invite an additional independent reviewer, or conduct further editorial assessment when additional expertise is required.

8. Revision and Re-review

When MLCS invites a revision, authors should submit a revised manuscript together with a point-by-point response explaining how they addressed each reviewer and editor comment. Authors should provide a reasoned explanation when they choose not to implement a requested change. The editor evaluates the revised manuscript and the authors' response. Manuscripts undergoing major revision may be returned to the original reviewers or sent to additional reviewers when necessary. The editor may assess minor revisions directly when further external review is not required. An invitation to revise does not guarantee eventual acceptance.

9. Editorial Independence and Manuscripts Submitted by Editors

MLCS applies the same scholarly standards to manuscripts submitted by members of its editorial team or editorial board. An editor who is an author, co-author, collaborator, or otherwise has a competing interest in a manuscript must not participate in its assessment, reviewer selection, correspondence, or editorial decision. The journal assigns such manuscripts to an independent editor who has no relevant conflict of interest. The submitting editor or editorial-board member does not receive privileged access to reviewer identities or confidential editorial deliberations. Editorial decisions are based on scholarly merit and must not be influenced by the authors' nationality, institutional affiliation, professional status, gender, beliefs, or other characteristics unrelated to the quality and integrity of the research. This approach is consistent with the principles of transparency and editorial independence promoted by COPE, DOAJ, OASPA, and WAME.

10. Ethical Concerns Identified During Review

If an editor or reviewer identifies potential plagiarism, duplicate publication, data fabrication or falsification, inappropriate image manipulation, unethical research practices, undisclosed competing interests, citation manipulation, or other possible misconduct, the journal may suspend the review while the concern is assessed. MLCS handles such cases according to its [Publication Ethics and Malpractice Statement] and relevant COPE guidance. Similarity-screening results do not automatically establish plagiarism. Editors assess the nature, extent, and context of any detected overlap before taking editorial action.

11. Supplementary Materials, Data, and Code

When supplementary files, datasets, code, protocols, or other materials form an essential part of the evidence supporting a manuscript, MLCS may make those materials available to reviewers as part of the peer-review process. Reviewers should evaluate such materials when they are necessary to assess reproducibility, methodological validity, or the interpretation of the findings. Authors remain responsible for ensuring that shared material complies with applicable requirements for privacy, informed consent, confidentiality, copyright, and data governance.

12. Appeals

Authors may appeal an editorial decision when they believe that a significant scientific or procedural error affected the assessment of their manuscript. An appeal must provide a clear scholarly justification and address the specific issues underlying the editorial decision. Disagreement with a decision alone does not constitute sufficient grounds for an appeal. The Editor-in-Chief reviews the appeal or assigns it to an independent editor when a conflict of interest exists. The journal may seek additional expert advice when necessary. An appeal does not guarantee reconsideration or acceptance of the manuscript.

Further procedures are described in the [Complaints and Appeals Policy].

13. Peer-Review Records

MLCS maintains appropriate records of editorial decisions, reviewer reports, author responses, and relevant correspondence in its journal management system in accordance with its editorial and data-management procedures. Reviewer identities and confidential review records are not disclosed publicly unless disclosure is required by law or necessary for the investigation of serious publication-integrity concerns.

14. Scope of This Policy

This peer-review process applies to Original Research Articles, Review Articles, and Short Communications published as scholarly contributions in MLCS. Any content that does not undergo the standard external peer-review process will be clearly identified as such on the published article.