Analysis of Ship Track Data Based on Multi Model Weighted Anomaly Detection
JIANG Zhihao
WEI Qiang
LI Shaomeng
ZHANG Dongning
Abstract：Aiming at the anomaly detection of ship track data,combining the advantages of neural network weight calculation and Boosting the idea of voting model,a multi-model weighted anomaly detection algorithm is proposed,which selects different types of unsupervised anomaly detection algorithms as the base model and trains feedforward neural network parameters with labeled data sets in different fields.The weight of each model is optimized by AUC score.By comparing the accuracy rate and processing speed of each model,the feasibility and correctness of multi-model weighted anomaly detection algorithm in ship track data analysis are verified.
Keywords：ship trackanomaly detectionmulti model weighteddata analysis
Publication Date：2025-11-20
Online Publishing Date：2026-01-28(First online date of this platform, not the publication date of the document)
Pages：5( 45-49 )