Optimization Research on Enterprise Human Resource Allocation via Hierarchical Clustering

This study focuses on 326 employees from IT company, employing the Hierarchical Clustering algorithm to conduct a clustering analysis of employees’ skill characteristics. The aim is to reveal the patterns of skill differences among employee groups and achieve scientific classification. During the research process, the silhouette coefficient was used to verify the rationality of the clustering results. It was found that when the number of clusters is 4, the silhouette coefficient reaches its maximum value, indicating that this clustering result has high intra-cluster similarity and significant inter-cluster differences. Based on the visual analysis of the clustering result distribution map, employees are divided into four groups with distinct skill characteristics: ‘innovative-driven’, ‘technical-analyst’, ‘executive leader’, and ‘communication-teamwork’. Further comparison between employees’ actual characteristics and clustering results through a heatmap confirms that the 4 clusters generated by the hierarchical clustering algorithm can effectively capture the essential characteristics of the preset types, realizing the macro-level merging of similar skill types. The research results clearly present the core competencies, scale distribution, and corresponding relationships with original types of the four groups, providing a scientific basis for the rational allocation of enterprise human resources.