KPI Prediction Algorithm Based on Phase Space Reconstruction and Support Vector Machine

The key performance indicators (KPI) system is of the complexity of setting and data processing. And its effectiveness of the application to the performance evaluation results is questioned in colleges and universities. With sufficiently analyzing on the characteristics of KPI data of colleges and universities, this article putts forward a more effective KPI prediction algorithm to achieve better data processing based on Phase Space Reconstruction and Support Vector Machine, and the pre-process of multi-filtering algorithm.