知识型员工心理契约预测方法研究--基于回归树和RBF神经网络的学习算法

摘要

With the psychological contract for the knowledgeable employees as the object, and the measured data from the re-search in three knowledgeable staff domestic enterprises, the predictive model of psychological contract based on RBF neural ne-twork is built. At the same time, an efficient and convenient regression tree combined with RBF neural network predictive model with the knowledgeable employees' psychological contract is established , in order to realize the optimization of RBF neural network prediction efficiency, select regression tree combined with RBF neural network, strive to achieve the complementary advantages of both. The results shows that through regression tree with RBF neural network prediction results compared with the actual measured data in three knowledge staff enterprises, the regression tree combined with RBF neural network prediction resu-lts has higher accuracy. Therefore, the learning algorithm of the regression tree-based RBF neural network and the prediction model for the algorithm-cored employees' psychological contract level prediction model are efficient.This model substitutes the pr-evious subjective measuring methods of psychological contract, meanwhile the establishment of the model provides managers with a scientific and efficient method for the forecast of the psychological contract for knowledgeable employee. The model to predict the level of the psychological contract for knowledgeable employee has higher accuracy.%以知识型员工心理契约为研究对象，结合在国内三家知识型员工企业调研的实测数据，建立基于RBF神经网络的心理契约预测模型，同时为实现对RBF神经网络预测效率的优化，选择回归树与RBF神经网络相结合的方式，力求实现两者的优势互补，建立一个高效便捷的回归树与RBF神经网络相结合的知识型员工心理契约预测模型。结果表明：通过回归树、 RBF神经网络预测数据结果与在三家知识型员工企业实际施测数据结果比较后发现，回归树与RBF结合的神经网络预测数据结果具有较高的准确性。因此，我们可以认为基于回归树的RBF神经网络的学习算法和以该算法为核心的知识型员工心理契约水平预测模型是有效的，该模型对知识型员工心理契约水平的预测具有较高的准确性。该模型能够替代以往对知识型员工心理契约主观预测的方法，使心理契约的预测过程更为高效，预测结果更加科学。