遗传算法在用户感知评估建模中的应用

[摘要] 在建立用户感知评估变量结构模型的基础上，利用遗传算法对KQI-QoE关系进行动态自适应建模，并与线性回归方程、指数回归方程、抛物线回归方程以及对数回归方程等传统的拟合方法进行了对比。实际用户感知和评估体系的拟合结果表明，遗传算法在用户感知建模方面具备良好的自适应搜索最优解的能力。这为进一步研究和完善用户感知评估体系提供了新的方法和途径。

[关键词] 移动；数据业务；用户感知；评估建模；遗传算法

[Abstract] In this paper, we use a variable structure model for evaluating customer experience and a genetic algorithm for dynamic adaptive modeling of KQI-QoE relations. We compare the genetic algorithm with traditional fitting methods, such as the linear regression equation, exponential regression equation, parabola regression equation, and logarithm regression equation. Actual customer experience and comparisons show that the genetic algorithm is good for searching for optimal solutions in terms of customer experience modeling. This is a new approach to improving the customer experience evaluation system.

[Keywords] mobile; data service; customer experience; evaluation model; genetic algorithm