Snatch theft crime for criminal patterns detection and classification using deep learning model / Nurul Farhana Mohamad Zamri

Every year, millions of crimes are reported all across the world. According to the statistical analysis of the crime rate for Malaysia, it shows that in Malaysia, the crime index ratio per 100,000 population was 273.8 cases in the year 2018. However, for WP Kuala Lumpur, for every 100,000 population it is 642.6 cases. Thus, it shows that crime usually happens within cities and towns. Besides the negative impacts on citizens' everyday lives, there is a significant impact on economic growth that shows the relationship between crime and economic growth in Malaysia. Hence, this study focused on snatch theft, including evaluation and validation in real-time detection, which has not been fully explored. This study aims to differentiate snatch theft scenarios from normal scenarios in predicting and detecting snatch theft crimes classification utilising snatch theft databases obtained from 120 videos on YouTube and Google.