Research and Application of LPRL Model Based on TOPSIS and K-Means Clustering

In this paper, LPRL evaluation model is established based on hierarchical analysis, entropy weight method and coefficient of variation method. First, according to the relevant criteria proposed by CIE, 15 sets of data were collected from 40 regions and integrated into 13 secondary indicators. We divided them into three dimensions, and in order to ensure the objectivity and accuracy of the weights, we used a combination of weighting method, combining hierarchical analysis, entropy weighting method and coefficient of variation method, to obtain the three-dimensional weights of these indicators. Then, the LPRL scores were calculated using the TOPSIS comprehensive evaluation method, and finally the classification was performed by cluster analysis. by calculating the LPRL scores, this paper uses the McKinsey matrix to measure the level. the classification range of LPRL and RDI is determined by the K-means clustering algorithm, and the McKinsey matrix is divided into nine quadrants, and the scores are reflected by the coordinates (LPRL, RDI) and the distance d. A new evaluation system model: LPRL2.0 was developed to provide some reference and meaning for other fields.