Back to Search
Start Over
Incremental classifier in crime prediction using bi-objective Particle Swarm Optimization
- Source :
- Information Sciences. 562:279-303
- Publication Year :
- 2021
- Publisher :
- Elsevier BV, 2021.
-
Abstract
- Nowadays, the increase in criminal activities has resulted in a massive generation of crime reports describing the details of the crime incidents. Analyzing these reports for crime type prediction helps the law enforcement agencies deal with crime prevention strategies. But it is quite a demanding and difficult task to consider these reports individually and determine their crime types. In the proposed work, an efficient classifier has been designed to analyze the crime reports which not only predict the crime types of the reports but at the same time upgrades itself with the help of new crime reports. Therefore, this task demands an incremental supervised learning technique that continuously learns the existing classifier based on the new set of reports and information already extracted from the old set of reports. Developing an incremental classifier infuses the knowledge that keep coming from the newly generated reports and help in increasing the report-discriminating power of the classifier. In this work, we have applied a Bi-objective Particle Swarm Optimization technique for generating an efficient incremental classifier for classifying and predicting the crime reports dynamically. Crime reports of different countries, such as India, the United States of America, and the United Arab Emirates, have been collected from online classified newspapers to measure the performance of the proposed as well as some state-of-the-art classifiers. Also, the method has been evaluated based on an unbiased police witness narrative crime reports and finally, a statistical test has been performed using all four considered datasets to measure the statistical significance of the proposed methodology.
- Subjects :
- Information Systems and Management
Computer science
ComputingMilieux_LEGALASPECTSOFCOMPUTING
02 engineering and technology
Machine learning
computer.software_genre
Theoretical Computer Science
Task (project management)
Set (abstract data type)
Artificial Intelligence
Crime prevention
0202 electrical engineering, electronic engineering, information engineering
Crime type
business.industry
05 social sciences
Law enforcement
050301 education
Particle swarm optimization
Computer Science Applications
Control and Systems Engineering
020201 artificial intelligence & image processing
Artificial intelligence
business
0503 education
computer
Classifier (UML)
Software
Subjects
Details
- ISSN :
- 00200255
- Volume :
- 562
- Database :
- OpenAIRE
- Journal :
- Information Sciences
- Accession number :
- edsair.doi.dedup.....29b7b9c1ecfea61afd12071d7d6542f6