Method to Predict and Explain Taste Preference using Pork Characteristics and Consumer Information.

As the demand for pork continues to increase, understanding consumer preferences for specific characteristics of pork has become increasingly important. Building on the insights provided by previous studies, this research aims to more accurately predict consumer taste preferences by utilizing both pork characteristic information and consumer information. To achieve this, the study proposes a deep learning-based stacking model that integrates the strengths of individual models, thereby preventing overfitting and improving prediction accuracy. Furthermore, this research introduces model explainability to enhance the understanding of the predictive results offered by the model, thus reinforcing the reliability and transparency of the outcomes.