Open AccessApplication of Evaluation Algorithm for Port Logistics Park Based on Pca-Svm Model
Author(s) -
Bianjiang Hu
Publication year - 2018
Publication title -
polish maritime research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.374
H-Index - 21
eISSN - 2083-7429
pISSN - 1233-2585
DOI - 10.2478/pomr-2018-0109
Subject(s) - port (circuit theory) , support vector machine , principal component analysis , computer science , scale (ratio) , data mining , operations research , algorithm , machine learning , engineering , artificial intelligence , physics , quantum mechanics , electrical engineering
To predict the logistics needs of the port, an evaluation algorithm for the port logistics park based on the PCASVM model was proposed. First, a quantitative indicator set for port logistics demand analysis was established. Then, based on the grey correlation analysis method, the specific indicator set of port logistics demand analysis was selected. The advantages of both principal component analysis and support vector machine algorithms were combined. The PCA-SVM model was constructed as a predictive model of the port logistics demand scale. The empirical analysis was conducted. Finally, from the perspective of the structure, demand, flow pattern and scale of port logistics demand, the future logistics demand of Shenzhen port was analysed. Through sensitivity analysis, the main influencing factors were found out, and the future development proposals of Shenzhen port were put forward. The results showed that the port throughput of Shenzhen City in 2016 was 21,328,200 tons. Compared with the previous year, it decreased by about 1.74 %. In summary, the PCA-SVM model accurately predicts the logistics needs of the port.