Parallel Animal Classification Analysis Based on Hierarchical Association Mining

  • Lingli Tan
Keywords: Animal Classification, Large Data, Association Mining, Parallel Clustering, Binary Tree, Support Vector Machine


In recent years, with the popularization of Internet of Things technology in the animal breeding industry,
the animal breeding industry has become more and more intelligent, large-scale, and standardized. In the
process of large-scale breeding of animals, timely classification of cultured animals can effectively avoid the infection
of immune diseases, and at the same time, the mature animals can be transported to the market in time to
avoid excessive breeding and effectively improve the efficiency of breeding. The purpose of this paper is to
achieve animal classification by correlating data mining methods. In order to improve the performance of animal
classification algorithm, a big data community discovery parallel clustering analysis method based on data
association hierarchy mining is proposed. Firstly, an evolutionary non-negative matrix factorization framework
based on clustering quality is proposed for dynamic community detection. Then, a clustering of dynamic pruning
binary tree SVM algorithm is proposed. The equivalence of evolutionary binary tree clustering and evolution
module density optimization is proved from the perspective of theoretical analysis. Based on this equivalence,
a new semi-supervised association mining algorithm is proposed by adding a priori information to the
sample data without increasing the time complexity. Finally, through the experimental analysis of static and dynamic
community detection models, the performance advantages of the algorithm for community detection performance
indicators are verified. The test results show that the method has significant effects on animal identification.