Modeling and predicting meat yield and growth performance using morphological features of narrow-clawed crayfish with...

Area of Science:

• Fisheries Science
• Aquaculture
• Computational Biology

Background:

• Traditional statistical methods struggle with complex, multi-parameter biological data.
• Artificial intelligence (AI) and machine learning (ML) offer advanced prediction capabilities for large datasets.
• Accurate modeling is crucial for sustainable management of fisheries and aquaculture resources.

Purpose of the Study:

• To predict length-weight relationships and meat productivity in narrow-clawed crayfish using ML models.
• To evaluate the performance of various ML algorithms for biological data analysis.
• To assess the utility of ML in fisheries and aquaculture management.

Main Methods:

• Utilized a dataset of 1416 narrow-clawed crayfish from Apolyont Lake, including morphometric measurements and growth performance.
• Applied seven different machine learning algorithms to predict length-weight relationships and length-meat yield for both sexes.
• Employed Support Vector Regression (SVR) and compared its accuracy against other ML models.

Main Results:

• Support Vector Regression (SVR) achieved high prediction accuracy, with values up to 0.996 for length-weight and 0.995 for length-meat yield.
• SVR demonstrated superior performance across all evaluated metrics, including accuracy, sensitivity, and specificity.
• The study validates ML and AI as effective alternatives to traditional estimation methods.

Conclusions:

• Machine learning, particularly SVR, provides highly accurate predictions for crayfish biological parameters.
• These findings support the integration of AI and ML in sustainable fisheries, aquaculture, and natural resource management.
• ML models offer a robust approach for future planning and decision-making in aquatic ecosystems.