TBESO-BP: an improved regression model for predicting subclinical mastitis

Area of Science:

• Veterinary Science
• Artificial Intelligence in Animal Health
• Dairy Production

Background:

• Subclinical mastitis significantly impacts dairy cow economics, welfare, and biosecurity.
• Current identification methods (SCC, microbiology) face challenges, limiting early intervention.
• Accurate prediction of subclinical mastitis is crucial for effective herd management.

Purpose of the Study:

• To develop and evaluate an enhanced neural backpropagation (BP) network model for predicting subclinical mastitis.
• To improve the accuracy and efficiency of somatic cell count (SCC) prediction using AI.
• To introduce the TBESO (Multi-strategy Boosted Snake Optimizer) algorithm for enhancing BP network performance.

Main Methods:

• Utilized monthly Dairy Herd Improvement (DHI) data from January to July 2022.
• Developed an enhanced BP network model incorporating the TBESO algorithm.
• Compared the TBESO-BP model against six alternative regression prediction models.

Main Results:

• The TBESO-BP model achieved a high coefficient of determination (R² = 0.94).
• The model demonstrated a Mean Absolute Error (MAE) of 2.07 and Root Mean Square Error (RMSE) of 5.33.
• TBESO-BP outperformed six alternative models in predictive accuracy and error reduction.

Conclusions:

• The TBESO-BP model is a precise and effective tool for predicting subclinical mastitis in dairy cows.
• The TBESO algorithm significantly enhances BP neural network performance for regression tasks.
• This AI-driven approach offers improved computational efficiency and practical application in dairy farming.