As the data about transportation of dangerous cargos on Yangtze River are incomplete, we constructed incomplete datasets for three types of data in the dangerous cargo transportation data on Yangtze River, i.e., the data of the number of ships registered for dangerous cargo transportation, the dangerous cargo carrying capacity, the dangerous cargo container carrying capacity, with a missing-data rate of 10% under the multi-variable missing-at-random mechanism. Then, we employed different interpolation methods to interpolate the missing data in the incomplete constructed datasets, comparing the performance of different interpolation methods by such indicators as the mean absolute error and the mean squared error. The research results revealed that the multiple interpolation (MI) method performed better than the expectation maximization interpolation (EM) and mean-value interpolation methods. The dataset obtained by multiple interpolation could well reflect the distribution pattern of the original data and the correlations between different types of data, which would be conducive to deep mining of data and optimization of the interpolation effect, providing a statistical basis for safety administration of dangerous cargo transportation on Yangtze River, facilitate design of precautionary measures against accidents and disasters, and improving the safety of waterborne transportation.
Application of interpolation method in data processing of dangerous cargo transportation in the Yangtze River
International Conference on Smart Transportation and City Engineering 2021 ; 2021 ; Chongqing,China
Proc. SPIE ; 12050
2021-11-10
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
British Library Conference Proceedings | 2021
|CHAPTER 6: DANGEROUS CARGO AND ``LEGALLY DANGEROUS'' CARGO
British Library Conference Proceedings | 2009
|Application of Modern Monitoring Measures in Lithuanian Dangerous Cargo Transportation
Springer Verlag | 2020
|The Yangtze River Transportation-Economic Belt
ASCE | 2008
|