Predicting Graduate Employment Quality in Agricultural Universities: A Machine Learning Framework Leveraging Multi-Dimensional 5-Point Likert Scale Survey Data

Breaking the talent bottleneck in agriculture and forestry and establishing an effective channel for transmitting intellectual achievements from university graduates to rural areas are crucial for building a high-quality rural revitalization workforce. This study employs a mixed-methods approach, combining systematic surveys based on a five-point Likert scale (Cronbach’s α = 0.982) with machine learning modeling to analyze the factors influencing the employment of graduates from agricultural and forestry institutions. Key findings indicate that respondents generally recognize the importance of salary and benefits, express high satisfaction with occupational environments and living conditions, and acknowledge the effectiveness of training systems and promotion channels. The Genetic Algorithm-Back Propagation (GA-BP) predictive model constructed in this study demonstrates outstanding performance, achieving coefficients of determination (R2) of 0.983 and 0.960 on the training and test sets, significantly outperforming traditional measurement methods. This research not only provides data-driven support for optimizing employment policies in agricultural and forestry institutions but also showcases an innovative application of artificial intelligence in analyzing employment factors, offering an interdisciplinary research paradigm for talent strategies aimed at advancing smart agriculture.