This document discusses a novel discrete version of the imperialist competitive algorithm (DICA) applied to the graph coloring problem (GCP), a well-known NP-hard optimization issue. The authors compare the performance of DICA against genetic algorithms (GA) using a series of benchmarks, demonstrating DICA's superiority in finding optimal and valid solutions. The paper details the theoretical foundation of GCP, the algorithms utilized, and presents experimental results that illustrate DICA's efficiency.