Comparison and Selection of Artificial Intelligence Technology in Predicting Milk Yield

Abstract

Forecasts are an effective decision-making tool, mainly in the dairy industry, because they help improve herd management, save farm energy and optimize long-term capital investment. The application of artificial intelligence technology to predict milk yield is a subject of concern in the scientific community. However, defining a technology or model to predict the effective performance of these products in different environments is a challenging and complex activity, because none of them is accurate in all scenarios. This study compared the application of artificial intelligence technology in milk yield prediction in the literature, and applied analytic hierarchy process to select the most suitable artificial intelligence technology for milk yield prediction. Methods comprehensive analysis, investigation and experiment were used. The results show that the artificial intelligence technology based on artificial neural network is more suitable for the prediction of milk yield than decision tree and support vector machine. In the field of milk production, the most relevant selection criteria are identified as the ability of these technologies to process uncertain data and their ability to obtain accurate results in the best way. The analysis carried out supports the decision-making of milk production organization.