An Adaptive Weighted Bagging Ensemble Learning Model for Zombie Enterprise Identification

Abstract

Zombie enterprise portrait and classification is one of the urgent problems in the current society, which has important practical significance and research value. This paper presents an adaptive weighted Bagging integrated learning method, which integrates 5 regular models and 8 pattern recognition models. The weight of the base classifier in integrated learning can be adjusted adaptively according to the training process to reduce the subjectivity and limitation of the regular model as much as possible. The accuracy, precision and recall rate of the model are all up to 1.0 in the experiment. At the same time, the strategy of data cleaning and missing item completion for problem domain is proposed. The research methods and results proposed in this paper have certain reference significance for the study of zombie enterprise portrait and classification and its related fields.