<![CDATA[The crime rate in Indonesia is highly increased. A lot of people want to become wealthy in a wrong way by commiting a crime. Criminality is an act that violates the rules of the law that can disturb the public. Every society has a risk of becoming a victim of crime. The greater the risk that the community has, the more unsafe their area is. However, the number of criminal acts cant't be ensured from time to time due to the uncertain number. This causes the police will having a trouble in resolving the criminal acts. A proper and accurate prediction can help minimizing criminal acts that will be happened. This research is intended to get predicted numbers of criminality using Extreme Learning Machine method (ELM). Based on the implementation and testing done by using crime data of Probolinggo District Police in 2012 until 2017, obtained the maximum network architecture that is the number of features as much as 7, the comparison of data ratio is 80%: 20%, and the number of neurons in the hidden layer as much as 7 and the binary sigmoid activation function. The low error value is calculated using the Mean Square Error (MSE) error rate and the result is 0.037662.]]>
Copyrights © 2018
