LSSVM Analysis of Competency Effects on Supply Chain Management from Individual and Collaborative Perspectives

In today's complex and ever-changing environment, concerns about having an excess of data have taken the place of those about insufficient data in supply chain management (SCM). The flood of data generated by each node in the supply chain has led to advancements in supply chain management analysis. The exponential expansion in data volume has rendered traditional methods inefficient and ineffective. Preprocessing, feature extraction, and model training are the three steps that make up the procedure. In data mining, data reduction is a crucial preprocessing step for generating a model that is accurate, rapid, and adaptable while also being characterized by low computational complexity. This will allow the model to respond quickly to incoming objects and changes. Designed to extract uncorrelated components from connected data, principal component analysis (PCA) is utilized in feature extraction. To train the model, the proposed approach utilized an LSSVM. With an average accuracy of 91.08%, the suggested method outperforms CNN and SVM.