<p>In machine learning, density-based clustering is a widely researched direction, and numerous methods have been proposed to address the challenge of clustering under different density conditions. These methods mainly classify samples by analyzing their local densities to achieve more accurate clustering results. However, for uneven density datasets, the limitations of local density estimation make it difficult to achieve effective clustering results. Therefore, this paper proposes a density clustering algorithm based on seed point with cluster expansion(DBCE). The method forms initial clusters from the seed point and dynamically adjusts the cluster expansion conditions according to the density distribution characteristics of the dataset. Then, we design a new proximity measure formula to assign the remaining points to the corresponding clusters finally. To validate the effectiveness of the method, we conduct experimental evaluations on both synthetic and real-world datasets. The results show that compared with the existing methods, the proposed method has higher clustering accuracy and computational efficiency for large-scale data.</p>