EUVIMEDEUVIMED entdecken
Crossref

FROM MORPHOLOGY TO MACHINE LEARNING: RANDOM FOREST-BASED APPROACH FOR FISH SPECIES IDENTIFICATION

Aalia Zafar, Muhammad Umer Sarwar, Muhammad Kashif Hanif, Ramzan Talib

The Journal of Animal and Plant Sciences · 13.07.2026 · 36 (5)

DOI: 10.36899/japs.2026.5.0121 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Journal of Animal and Plant Sciences, online erschienen am 13.07.2026. Verfasst von Aalia Zafar und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Fish are important components of freshwater ecosystems and are one of the main sources of protein for humans. One of the main problems in fishery resource investigation is the accurate identification of fish species and the accurate discrimination of fish stocks. In the context of large data, machine learning techniques as developing data processing techniques have gradually replaced traditional methods. This study demonstrated an automated study, using machine learning, to classify fish species by morphological features (e.g., body shape, size, and fins etc.). Machine learning-based fish species identification is essential for automated ecological monitoring, biodiversity assessment, and fisheries management. This study used a dataset of 540 samples across 20 distinct fish species, a Random Forest model was trained and evaluated using a 10-fold cross-validation strategy, outperforming alternative comparative models in classification accuracy. The procedure used to prepare the dataset consisted of preprocessing, including categorical to binary encoded categorical data (converting non-numeric data to numbers the machine could interpret). Several machine learning models were examined, with the Random Forest (RF) providing the best results, and the RF became a trained, tested, and saved machine learning model for user applications. Subsequently, a slight implementation interface was additionally developed to demonstrate the practical applicability of the trained classification framework that allowed users to enter fish features and receive predictions. The proposed framework illustrates how machine learning approaches may support species identification in fisheries and ecological studies, increase prediction accuracy, and have potential applications in fish farming and research as well as in aquatic ecosystem protection and conservation.

Zitieren

APA

Zafar, A., Sarwar, M. U., Hanif, M. K., & Talib, R. (2026). FROM MORPHOLOGY TO MACHINE LEARNING: RANDOM FOREST-BASED APPROACH FOR FISH SPECIES IDENTIFICATION. The Journal of Animal and Plant Sciences, 36(5). https://doi.org/10.36899/japs.2026.5.0121

Vancouver

Zafar A, Sarwar MU, Hanif MK, Talib R. FROM MORPHOLOGY TO MACHINE LEARNING: RANDOM FOREST-BASED APPROACH FOR FISH SPECIES IDENTIFICATION. The Journal of Animal and Plant Sciences. 2026;36(5). doi:10.36899/japs.2026.5.0121

BibTeX
@article{Zafar2026,
  author = {Zafar, Aalia and Sarwar, Muhammad Umer and Hanif, Muhammad Kashif and Talib, Ramzan},
  title = {FROM MORPHOLOGY TO MACHINE LEARNING: RANDOM FOREST-BASED APPROACH FOR FISH SPECIES IDENTIFICATION},
  journal = {The Journal of Animal and Plant Sciences},
  year = {2026},
  volume = {36},
  number = {5},
  doi = {10.36899/japs.2026.5.0121}
}
Kostenlos starten →