人工神经网络模型在研究生招生数量预测中的应用

Application of Artificial Neural Network in Predicting the Number of Graduate Students

研究生招生数量的确定涉国家政策、社会就业、人才需求、专业分布与需求等诸多因素,这些影响因素往往无法量化,而且各个影响因素之间关系错综复杂,简单的线性模型预测未来招生数量往往难以实现.尝试采用人工神经网络(BP-ANN)模型,针对历年招生数量原始数据信息零散、隐含影响因素过多、诸多影响因素难以确定性描述等问题,通过对黑龙江省历年(1981—2004)研究生招生数量进行系统分析,建立了人工神经网络预测模型,并对未来3年的招生数量进行了预测,预测结果较好,为该方面研究提供了新的研究思路与研究方法.

Determination of the number of graduate students involves in many factors, including national policy, employment, the demand of human resources, professional distribution and requirement, etc. These influencing factors can not be quantified and the relationships between them are complex. It is difficult to predict the future number of students with simple linear model. In this paper, aiming at the problems that raw data information of the number of students is scattered in the past few years, too many influencing factors are hidden and many influencing factors are difficult to be certainly described, we have tried to use artificial neural network （BP-ANN） model, and constituted artificial neural network prediction model by systematically analyzing the number of graduate students over the past few years in Heilongjiang Province, then we have also predicted the number of students for the next 3 years. Better predicting results have provided new research ideas and methods for the study in th is field.