استخدام الشبكات العصبية الاصطناعية في التنبؤ بمعدلات الجريمة في قطاع غزة

Abstract EN

It has become recognized the importance of security planning in the security services, especially in the crime combat field , In order to success the security planning process , it should be based on accurate criminal statistics and information , The process of crime predicting , its rates and its indicators is very important part of criminal statistics process , It contributes significantly to support the security plan in process of developing appropriate plans to combat crime.

So This study examined the use of artificial neural networks in predicting the crime rates in the Gaza Strip , where they were dealing with the crude rate of crime in the Gaza Strip per 100,000 inhabitants , which has a value of atypical , was the use of models of artificial neural networks (ANN) in two cases , namely the case of a value anomalous , in the case of treatment value anomalies , and the comparison between the models in terms of prediction accuracy and efficiency , has been the trade-off between the estimated models through the use of a standard trade-off (AIC, BIC), and I study concluded that models of neural networks (ANN) after processing the value anomalies were the most efficient The highest predictive ability models of neural networks (ANN) without addressing the value of the anomalous.