文章针对我国某类研究学者的年薪进行了调查研究。调查表明,工薪阶层的年薪主要与他们的研究成果的质量指标、从事研究工作时间以及成功获得资助的指标等有主要关系。在调查数据基础上,应用人工神经网络理论,建立我国某类研究学者年薪预测模型。通过验证表明,经过训练的年薪预测神经网络模型预测效果较好。该模型为国家和用人单位合理地安排预算以及提升自身竞争力具有一定的应用价值和实际意义。 The article surveys the annual salary of a certain kind of research scholars in our country. The survey shows that the salaries of working-class people are mainly related to the quality indicators of their research achievements, the time spent in research work and the indicators of successful funding. Based on the survey data, the artificial neural network theory is applied to establish the annual salary prediction model of some Chinese scholars. The verification shows that the trained annual salary prediction neural network model is better. The model has certain practical value and practical significance for the country and the employing unit to reasonably arrange the budget and enhance their own competitiveness.