Applying Levenberg Marquardt Algorithm in Feedforward Neural Network Models for Predicting Crime in Public Management
Artificial intelligence applications have been tremendously increased in various science fields, the last decades, with the development of new machine learning techniques and algorithms and of new artificial neural network tools for developing neural network models. In this research, the application of Levenberg Marquardt algorithm in Feedforward Neural Network Models for predicting crime urban data is implemented. The Levenberg Marquardt algorithm is a combination of the steepest descent algorithm and the Gauss-Newton algorithm which is used for solving non-linear least-squares problems. The algorithm combines the minimization advantages of the steepest descent method with the quadratic model of Gauss-Newton method in order to increase the speed of the overall process of finding the minimum of a function. Multilayer Feedforward Perceptron was utilized as it is considered as the most suitable for time series predictions, among several training algorithms such as backpropagation. Urban crime forecasting can play a significant role in urban planning and public management by facilitating decision making and the adoption of the most adequate proactive strategies in crime prevention and in public safety management planning.