Clarifying the features of the packaging of cultural and creative products is beneficial to out-of-the-loop cultural and creative products. In this paper, starting from probability theory, graph theory and information theory principles, we combine Bayesian network structure and feature selection theory to construct a feature extraction model for cultural and creative product packaging. The maximum mutual information principle is used to optimize the feature selection scoring function, and experimental analyses of execution time and performance comparisons are conducted for the BNCC-FS method. In terms of execution time, the average training time of the BNCC-FS method on the eight datasets is 4.73s, which is 14.53% and 43.44% lower compared to CC-FS and ECC-FS, respectively. In terms of performance comparison, the average HammingLoss value of the BNCC-FS method on sixteen datasets was 0.155, which was 2.52% and 10.41% lower compared to CC-FS and ECC-FS, respectively. This shows that the Bayesian network-based feature extraction method can effectively achieve feature extraction of the data and also provides a new research database for the analysis of the packaging features of cultural and creative products.
Clarifying the features of the packaging of cultural and creative products is beneficial to out-of-the-loop cultural and creative products. In this paper, starting from probability theory, graph theory and information theory principles, we combine Bayesian network structure and feature selection theory to construct a feature extraction model for cultural and creative product packaging. The maximum mutual information principle is used to optimize the feature selection scoring function, and experimental analyses of execution time and performance comparisons are conducted for the BNCC-FS method. In terms of execution time, the average training time of the BNCC-FS method on the eight datasets is 4.73s, which is 14.53% and 43.44% lower compared to CC-FS and ECC-FS, respectively. In terms of performance comparison, the average HammingLoss value of the BNCC-FS method on sixteen datasets was 0.155, which was 2.52% and 10.41% lower compared to CC-FS and ECC-FS, respectively. This shows that the Bayesian network-based feature extraction method can effectively achieve feature extraction of the data and also provides a new research database for the analysis of the packaging features of cultural and creative products.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.