A survey on crime analysis and prediction. (2022)

• Record Type: Journal Article
• Authors: Thomas, Ashly; Sobhana, N.V.
• Is Part Of: Materials today. Volume 58:Part 1(2022)
• Page Start: 310
• Page End: 315
• Publication Date: 2022
• Subjects: Deeplearning -- Neural networks -- Spatio-temporal -- Regression; Materials science -- Congresses -- Periodicals; 620.1
• DOI: 10.1016/j.matpr.2022.02.170
• Languages: English
• ISSNs: 2214-7853

Abstract

Abstract: A lot of police forces around the world have adopted mechanisms that use statistical data to guide their decision-making which is coined as predictive policing. As urbanization is increasing day by day, it is highly demanding to keep an eye on the criminal activities of each region so as to reduce the occurrences of unwanted behaviours. Prediction of crimes can be done only using the analysis of the patterns of criminal activities using the past data available with the concerned personals. This mainly makes use of the historic data and analyze them using Deep learning, Statistical Models and Algorithms. This paper makes a study on different approaches used worldwide for the prediction and forecast of crime occurrences. The methods are categorized and their effectiveness in various areas based on the precision and accuracy in their prediction is studied so as to show a light to the existing methodologies and to the need for future developments.