符号网络预测准确度及时间代价的优化

Research about Optimizing Prediction Accuracy and Time Complexity in Signed Networks

符号网络的预测准确度越来越高,但是时间复杂度也越来越难以接受。必须寻找有效预测方法,既保证算法预测准确度高,同时时间复杂度低。本文设计了一个优化算法,使用平衡环算法预测符号,利用函数拟合方法分别拟合预测准确度与步长、时间复杂度与步长的函数关系,分析随步长增加预测准确度与时间复杂度的关系并提出优化方案。实验显示,本文的优化算法能够有效获得预测准确度与时间复杂度的关系。本文可供设计符号预测算法的研究者参考。

In signed networks, different sign predicting algorithms have been proposed. The prediction accuracy of the algorithm is improving, but the time complexity is also increasing. A way must be found to reduce the time complexity. In order to ensure the high predict.ion accuracy and low time complexity, an optimization algorithm is designed to analyze the relation between prediction accuracy and time complexity with increasing steps and an optimization scheme is also proposed through using the balanced cycle algorithm for predicting sign at first and then fitting the function of prediction accuracy and step, time complexity and step respectively. Experiments show that the optimization algorithm can effectively obtain the relation between prediction accuracy and time complexity. This research can be used in working out design symbol prediction algorithms.