Research and Application Based on IRA Quantitative Model and K-means Clustering

This study constructs a comprehensive adaptation model system, which consists of three core models: The IRA model quantifies the grades through four types of indicators; the REV model introduces indicators to optimize the decision-making process; and the BCA model formulates priority plans by adopting a combination of subjective and objective weighting methods. In the construction of the models, the entropy weight method, as an objective weighting method, is widely applied. Its core idea is to determine the weights by calculating the information entropy of each indicator. In addition, the K-means clustering algorithm is also applied in this study for classification purposes. The K-means clustering is a classic unsupervised learning algorithm. It divides the data into K clusters by iteratively optimizing the position of the centroids. In this study, the K-means clustering algorithm is used for classification. By dividing the data into different clusters, the characteristics can be identified more clearly, thus providing a basis for formulating targeted strategies.