Predicting age at first calving of dairy breed calves using whale optimization-based ensemble learning framework.

Area of Science:

• Animal Science
• Data Science
• Machine Learning

Background:

• Ethiopia faces a growing dairy demand-supply gap due to feed shortages, high costs, and poor local breed genetics.
• Modern breeding facilities and advanced technologies like big data analysis and machine learning are crucial for addressing these challenges.

Purpose of the Study:

• To develop a predictive model for age at first calving in weaned calves.
• To utilize pre-weaning and weaning parameters for accurate age prediction.

Main Methods:

• Development of predictive models using Support Vector Regression (SVR), Linear SVR (LSVR), and Nu SVR.
• Hyperparameter tuning of SVR models, achieving 96.46% accuracy.
• Feature optimization using Whale optimization technique and ensemble modeling of SVR, LSVR, and NuSVR.

Main Results:

• The optimized SVR model achieved 96.46% accuracy.
• The ensemble model, trained on optimized features, reached a superior accuracy of 98.3%.
• The developed model effectively predicts age at first calving using calf and dam parameters.

Conclusions:

• Machine learning, specifically ensemble SVR, offers a powerful tool for predicting age at first calving in dairy calves.
• Accurate prediction can aid in optimizing breeding strategies and improving dairy production efficiency in Ethiopia.
• This approach can contribute to mitigating the milk supply deficit by enhancing livestock management.