I. Introduction
The graph coloring problem has always been a classic topic in graph theory [1], with significant value in theoretical research and practical applications. In recent years, with the vigorous development of information technology, it has strongly promoted the progress of graph theory. Many scholars have conducted a series of research and exploration around it, leading to the introduction of numerous new concepts, such as vertex-distinguishing total coloring [2]–[3], edge-magic total labeling, and vertex-magic edge coloring, among others. In addition, with the development of machine learning technology, scholars have also attempted to use techniques such as neural networks [4]–[5] to solve the graph coloring problem. Although still in the exploratory stage, it holds great research prospects.