中文摘要：

随着多特征决策研究的深入，传统方法已经不能回答更加细致的问题。细察精确预测的理论、建立模型与数据的形式关系成为更有希望的研究方向。神经网络模型设计用来模拟许多并行的认知和神经行为，具有样例学习和迁移适应能力．在解释和预测方面具有传统方法所不具备的潜力。神经网络能够同时表征线性补偿和非补偿规则，其应用已经渗透到许多学科领域。网络范式对于人事研究和应用也有价值，有研究表明神经网络可以用于人力资源管理的一般领域，成为人事决策研究的新范式。

英文摘要：

There has been tremendous progress in the investigation of multi - attribute decision making during the last four decades. The old methods may no longer suffice to answer more detailed questions. Scrutinizing the theories for precise predictions and formulating formal relationship between models and data might be a promising way to grasp those questions. Neural networks can be thought of as computer models designed to recognize meaningful patterns in data. Artificial neural networks are much better in predictive accuracy than conventional statistics, The neural network can represent both linear compensatory and noncompensatory hales exactly, Suecessful neural network applications in other disciplines suggest that the network paradigm may also be useful for personnel researchers and practitioners.