Estimation of the Length at First Maturity of the Swimming Crab (Portunus trituberculatus) in the Yellow Sea of Korea Using Machine Learning

  1. Jaehyung Kim
  2. Daehyeon Kwon
  3. Soojeong Lee

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Swimming crab (Portunus trituberculatus) is a commercially valuable species in the Yellow Sea, where recent fluctuations in resource levels have raised concerns about sustainable management. This study aimed to improve the estimation of the carapace length at 50% maturity (L50) using machine learning techniques, providing a more consistent and reproducible framework for visual maturity classification by standardizing image-based decision processes. Using geometric image augmentation (e.g., rotation, flipping, brightness adjustment), Hue–Saturation–Value (HSV) color segmentation, and algorithms, such as Extreme Gradient Boosting (XGB), Support Vector Machine (SVM), Random Forest (RF), and ensemble models, we classified the maturity of female crabs based on gonad color features. Model performance was evaluated using accuracy, AUC, and the TSS, with the ensemble model showing the highest predictive capability. The machine learning-based L50 was estimated at 64.63 mm (±1.73 mm), yielding a narrower uncertainty range than the visually derived L50 of 65.47 mm (±2.89 mm) under the same macroscopic labeling framework. These results suggest that machine learning-assisted maturity classification can enhance the precision and operational consistency of maturity estimation under a standardized framework, while biological accuracy cannot be confirmed in the absence of an independent reference, such as histological validation.

Article activity feed

  1. Version published to 10.3390/jmse14040335
  2. PREreview

    This Zenodo record is a permanently preserved version of a PREreview. You can view the complete PREreview at https://prereview.org/reviews/18281193.

    The title of this preprint on PREreview is funny, it contains a lot of meaningless wordings. Excluding those funny words, the content of the manuscript is about machine learning, we are not sure if the machine learning devices or applications caused such errors. Or else, it maybe a language translation error of the system.

    <span class="word">Estimation <span class="word">of <span class="word">the <span class="word"><span class="changedDisabled">First <span class="word"><span class="changedDisabled">Maturity <span class="word"><span class="changedDisabled">Using <span class="word"><span class="changedDisabled">Machine <span class="word"><span class="changedDisabled">Learning <span class="word">of <span class="word"><span class="changedDisabled">Swimming <span class="word"><span class="changedDisabled">Crab (<em><span class="word italic">Portunus <span class="word italic">trituberculatus</em>) <span class="word">in <span class="word">the <span class="word">Yellow <span class="word">Sea <span class="word">of <span class="word">Korea

    The true title should be:

    Estimation of the First Maturity Using Machine Learning of Swimming Crab (Portunus trituberculatus) in the Yellow Sea of Korea

    From our limited information technology knowledge, such discrepancies in the displayed title likely stem from the underlying HTML markup used by the platform (possibly a preprint aggregator like Scilit or a similar site, which aggregates content from sources like Preprints.org). These tags with classes like "word" and "changedDisabled" are not part of the actual manuscript title but are added in the web page's source code for technical purposes, such as:

    • Change tracking or diff visualization: The "changedDisabled" class suggests a system for marking edited or revised text (e.g., from document versions or collaborative edits), where changes are noted but not actively shown or "disabled" in the final view. This could occur if the title was parsed from a document with track changes enabled, like in Microsoft Word or a content management system, and the platform retains the markup during aggregation.

    • Word-level processing: Wrapping each word in allows for features like search term highlighting, accessibility enhancements (e.g., for screen readers), or interactive elements (e.g., tooltips, annotations, or JavaScript-driven behaviors in review platforms).

    • Parsing or import artifacts: When pulling metadata from the original preprint server (e.g., Preprints.org), the aggregator might introduce this structure during HTML conversion, especially if the source involves formatted text like italics for scientific names, which can lead to nested tags.

    In a browser, these tags are invisible, and the title renders cleanly as plain text. The "discrepancy" becomes apparent only when viewing the page source, inspecting elements in developer tools, or copying the content in a way that includes the raw HTML (e.g., via certain browser extensions or feeds). If the platform has a bug or incomplete sanitization, the tags might occasionally display as text, but that's less common. The true title is the unadorned version without this markup, as it's the canonical text from the preprint.

    Competing interests

    The authors declare that they have no competing interests.

    Use of Artificial Intelligence (AI)

    The authors declare that they did not use generative AI to come up with new ideas for their review.

  3. Version published to 10.20944/preprints202601.0984.v1