AN Identification and Prediction Model Based on PSO

K-Means Clustering Model

The K-means clustering model is a classic unsupervised clustering model and one of the top ten pattern recognition models (Wang et al., 2007; Yu et al., 2007; Deng et al., 2016; Zhao et al., 2022). Because the idea and implementation of the K-means clustering model are relatively simple, and the clustering accuracy is high, it is called a clustering model with good practicability (Wang et al., 2012; Fan et al., 2021; Subramaniyaswamy et al., 2017). The main idea of the K-means clustering model is to divide the samples in the sample set into K clusters according to the distance between the samples in the sample set (Huang et al., 2016; Sha’abani et al., 2020; Chen et al., 2009, 2020). Make the points in the clusters as close together as possible and make the distance between the clusters as large as possible (Alkhatib et al., 2013). From another perspective, the K-means clustering model quantifies the requested data into clustering centers. The objective function in the K-means clustering model is to minimize the data loss during the quantization process. The process of -means clustering model to solve the objective function can be regarded as expectation maximization (EM) iterative optimization.