Computo relies on Open Review for the review process. The review form is text-based, but Markdown and LaTeX formatting is supported so you can add hyperlinks and use LaTeX to add equations to your review. Reviewers are also required to answer a handful of rating scale questions about the submission.
Once a manuscript is accepted, reviews and discussion will be made available on the Computo website. Reviewers can choose to remain anonymous or not.
Guidelines for evaluation
In order to help you in performing your review we provide a list of the main questions we are trying to answer when evaluating a submission:
Is the paper within the scope of Computo?
See Aims and Scope of Computo.
Is the paper clearly written?
Computo is intended for computational scientists in statistics/machine learning. The Abstract and Introduction should be as nontechnical as possible, and provide a clear description of the contributions of the paper. Strength and limitations of the work should be adequately discussed, in particular in relation to related work. Graphs and tables should be well thought out and uncluttered.
Is the paper correct?
Mathematical and algorithmic validity are the authors’ professional responsibility. Referees can spot errors of reasoning, but are not expected to perform line-by-line checks of technical results.
Is the paper adequately evaluated?
Are all claims clearly articulated and supported either by empirical experiments or theoretical analyses? If appropriate, have the authors implemented their work and demonstrated its utility on a significant problem?
Is the paper reproducible?
The reproducibility of numerical results is a necessary condition for publication in Computo. The referees are expected to check whether they can run the code provided by the authors to reproduce their results. In case of major reproducibility issues, the referees should warn the Associate Editor as soon as possible.
Computo abides by the Committee on Publishing Ethics’s (COPE) guidelines listed below on plagiarism: