DATA MINING TECHNIQUES FOR THE AUTOMATIC CLASSIFICATION OF SHIP TRACKS

PDF

Authors
  1. Ali, A.S.
  2. Gelinas, J.
Corporate Authors
Defence Research Establishment Valcartier, Valcartier QUE (CAN)
Abstract
This technical note documents the work done during the summer 1998 as part of the ongoing project "Data Mining Techniques for the Exploitation of Archived Information". Raw data collected from a coastal surveillance radar was analysed in order to evaluate the possibilty of using it to automatically classify ships navigating in the open seas near the Canadian Coasts. The results based on data samples show that the fishing vessels could be separated from other ships with a near perfect accuracy. Another possible classification of ships into normal and abnormal (not fitting any reasonable pattern) was also attempted successfully. It is suggested that this tudy be continued using more data and better attributes to represent the data.
Keywords
Data mining;Ship tracks;Ship classification;Supervised learning;Machine learning;Classifiers
Report Number
DREV-TN-9809 — Technial Note
Date of publication
01 Nov 1998
Number of Pages
55
DSTKIM No
98-02484
CANDIS No
509431
Format(s):
Hardcopy;Document Image stored on Optical Disk

Permanent link

Document 1 of 1

Date modified: