Performance Prediction and Analysis of Regional Breeding Pigs Based on Integrated Learning—Take Yibin Academy of Agricultural Sciences as an Example

Taking Yibin Academy of Agricultural Sciences as an example, this study discusses the performance prediction and analysis method of regional breeding pigs based on integrated learning. Firstly, a complete data set was constructed by collecting and sorting out a large number of performance data of breeding pigs provided by Yibin Academy of Agricultural Sciences. Then, the ensemble learning method, including decision tree (DT), random forest (RF), support vector machine (SVM) and neural network, is used to predict and analyze the growth and development of breeding pigs. The experimental results show that compared with the single model, the ensemble learning model has significantly improved the accuracy, recall and F1 value, which proves the effectiveness and superiority of the ensemble learning method in the performance prediction of breeding pigs. The effects of different voting strategies on the performance of the ensemble learning model are further compared. The results show that the exponential weighted voting strategy is the best in all performance indicators. Finally, through the analysis of learning curve and verification curve, it is found that the integrated learning model has good training process and generalization ability. This study provides farmers with more scientific and reliable breeding guidance and decision support, and also provides a new empirical research case for the application of integrated learning method in agriculture.