AR 06-03
A Cargo Security Early Warning System - The Application of Neural Networks to Detect Cargoes with Potential Security Fraud
Melody Kiang
Department of Information Systems
College of Business Administration
California State University, Long Beach, CA 90840
Tel: (562)985-8944
Fax: (562)985-5543
Email: mkiang@csulb.edu
Robert Chi
Department of Information Systems
College of Business Administration
California State University, Long Beach, CA 90840
Tel: (562) 985-4238
Fax: (562)985-5543
Email: rchi@csulb.edu
Project Objectives:
The objective of this study is to explore the possibility of applying neural networks as an early warning system to alert port authorities with cargoes of potential security fraud. In other words, when fed the appropriate data inputs regarding cargo containers, this system would "learn" the differences between potentially lawful and unlawful cargoes. Unlawful cargo would include forms of contraband such as untaxed cigarettes, infested fruit, counterfeit software, illegal immigrants, narcotics, drugs, "dirty bombs" (i.e., explosives filled with nuclear waste), weapons, and other terrorist devices. X-ray, radiation scanners, and cameras can only go so far as to detect these forms of contraband. A neural networks based early warning system will allow for port personnel to focus their investigation efforts on cargoes that are more likely to contain contraband, which helps to improve detection efficiency and safety.
Tasks:
1. Preliminary Study and Data Identification
2. Data Collection and Analysis
3. Train and Test the System
4. Result Validation and Delivery of Final Report
Milestones/Dates:
7/1/06 through 6/30/07
Total Budget:
$40,000
Student Involvement:
Student Assistants: 406 hours at $12/hr
