Prediction and Empirical Research on Matching Values of Person-Job Measurement Based on EfficientNet

Abstract

In today's social environment, person-job matching is of significant importance for enhancing employee satisfaction, reducing turnover rate and improving organizational performance. However, traditional job matching methods rely on manual assessment and questionnaire surveys, which not only consume a lot of time and energy, but are also susceptible to the influence of subjective factors, resulting in a significant reduction in the accuracy of the assessment results. In order to overcome these challenges, the study proposes an EfficientNet-based job matching prediction model. After 300 epochs of iterations, the highest classification accuracy of this model can reach 0.844, indicating that the model has excellent data classification ability. When the quantity of training samples is increased to 16000, the accuracy of the model prediction increases and reaches 0.844, and the fitting error is reduced to 0.413. By adaptively adjusting the network structure and parameters, the model significantly improves its performance while keeping its size constant. The study shows that the model improves the accuracy and efficiency of person-job matching, which has important research and application value for modern organizational management.

翻译

搜索

复制