Cluster Analysis of Talent Portraits Based on K-means Algorithm

With the rapid development of power communication infrastructure and the further refinement and precision of power system business information, the power system platform generates and stores a large amount of talent information and business data in the management process. Human resource management has become a critical task. Due to many ideological and technical limitations in the past, talent management methods can't meet the talent management needs of today's colossal power system, which seriously affects the promotion, application, popularization, and development of the talent management model in the new era. Therefore, this paper is mainly based on the demand response research of talent portrait characteristics of power companies. Firstly, this paper uses data mining theory and clustering analysis method to extract and analyze the occupational characteristics of employees in power companies and discusses the clustering application of the K-means algorithm in high-dimensional data such as talents, skills, and technical majors. Then, on this basis, the improved K-means algorithm is introduced, the improved K-means algorithm is put forward, and its effectiveness in talent occupation clustering is verified. Finally, through the cluster analysis of talents' occupations in electric power companies, this paper extracts and classifies the characteristics of talents' professional abilities and puts forward a talent-side interactive response strategy method based on the characteristics of talents' portraits in electric power companies.