Ge, L.; Li, J., and Chen, J., 2020. Research on seafood traceable data based on k-modes clustering algorithm. In: Yang, D.F. and Wang, H. (eds.), Recent Advances in Marine Geology and Environmental Oceanography. Journal of Coastal Research, Special Issue No. 108, pp. 73–77. Coconut Creek (Florida), ISSN 0749-0208.
The improved k-modes clustering algorithm for classification attribute data is proposed, in which the classification attribute data of seafood traceable data are taken as an example and the k-modes algorithm is improved by improving clustering accuracy and process. The improved k-modes clustering algorithm combines density and distance to select the initial center point of cluster, which ensures the effectiveness of the initial cluster center and avoids falling into local extreme points so that clustering process can be simplified. The improved k-modes clustering algorithm redefines the clustering mode and comprehensively considers the representative of all attribute values in sample attributes to clustering category so that the distance measure will be improved and that the clustering effect can be optimized. Experiments show that the improved k-modes clustering method has good clustering effect on standard data sets, which is universal to classification attribute data and worthy of being more popular and further improving in practice.