基于KPCA算法的高校大学生网络消费预警研究

Research on Early Warning of College Students Online Spending based on KPCA Algorithm

当今各高校的学生网络消费趋势逐渐加强,为促进高校大学生形成良好且积极的消费观念并进行合理评价与校正,通过多渠道收集了在校大学生的月消费总额以及各个消费领域的金额。采用KPCA算法,建立了针对高校大学生的网络消费预警模型,通过实例分析建立了消费回归曲线,将结果与PCA算法进行对比。结果表明:该模型较PCA算法表现出更显著的代表性,消费回归曲线显示对原数据具有良好的拟合效果,算法所确立的函数预警阈值能对不合理的消费行为进行有效预警,这些均对促进大学生形成健康的消费观念具有积极作用。

Online spending of college students has been increasing, and it is necessary to guide them to develop healthy spending habit and value. This essay has collected the data of monthly online spending of colleges students, uses KPCA algorithm to construct an early warning model, establishes the consumption regression curve through example analysis, and compares the results with PCA algorithm. The results show that the model is more representative than PCA algorithm, the consumption regression curve shows a good fitting effect on the original data, and the function warning threshold established by the algorithm can effectively warn unreasonable consumption behavior, which plays a positive role in helping college students to develop a healthy consumption concept.