← Back to Publications List

Genetics and Breeding Approaches for Disease-Resilient and High-Yield Fish Species in Bangladesh

Students & Supervisors

Student Authors
Ekramul Hasib
Bachelor of Science in Computer Science & Engineering, FST
Abrar Sharif
Bachelor of Science in Computer Science & Engineering, FST
Sakib Rahman
Bachelor of Science in Computer Science & Engineering, FST
Supervisors
Md.mortuza Ahmmed
Associate Professor, Faculty, FST

Abstract

The present study proposes a data-centric strategy that targets fish breeding and genetic selection to obtain the main goal of making populations of fish that can resist diseases and are of high yield. A substantial dataset consisting of biological, environmental, and nutritional aspects was put together to create an effective representation of fish health and growth conditions. Different machine learning methods such as Linear Regression, Random Forest, XGBoost, Gradient Boosting, and AdaBoost were trained and assessed to forecast the Health Impact Score of different fish species. The application of Explainable Artificial Intelligence (XAI) techniques, including SHAP, LIME, and ELI5, to the intersected models helped not only to visualize feature importance but also to uncover the underlying decision-making process of the models. Among all models tested, the model that stood out in terms of its predictions and its capability to deal with complex and non-linear variables relevant to fish health prediction is Gradient Boosting, getting an accuracy rating of 94.8%. XGBoost is ranked second, getting an accuracy rating of 92.6%, proving to be somewhat less accurate than Gradient Boosting. Overall, the models performed in the following manner: Gradient Boosting (94.8%) > XGBoost (92.6%) > Random Forest (91.3%) > AdaBoost (89.7%) > Linear Regression. The proposed framework thus integrates predictive analytics and insights with explainability, providing proof for the concept of using data-driven decision-making principles for sustainable and intelligent aquaculture systems. The results also emphasize the importance of combining Artificial Intelligence and Explainable Artificial Intelligence for the development of precision fish breeding or environmentally sustainable and friendly aquaculture systems.

Keywords

Fish Breeding Genetic Selection Machine Learning Explainable AI Gradient Boosting SHAP LIME ELI5 Sustainable Aquaculture.

Publication Details

  • Type of Publication:
  • Conference Name: 1st International Conference on Life Science, Health, and Biotechnology (LifeTech 2026)
  • Date of Conference: 17/01/2026 - 17/01/2026
  • Venue: Jashore University of Science and Technology, Jashore 7408, Bangladesh
  • Organizer: Jashore University of Science and Technology
View on Publisher