Data-driven models in machine learning for crime prediction
Zbigniew Maciej Wawrzyniak, S. Jankowski, Eliza Szczechla, Zbigniew Szymanski, R. Pytlak, Paweł Michalak, Grzegorz Borowik · 2018
Prediction of future events over time is associated with a sequence of time series observational samples and other exogenous data. Different approaches connected with statistical learning techniques result in predictive data-based models. The paper presents an attempt to develop techniques for predictive data-based modeling based on machine learning data-driven approaches. To reach a good level of prediction we use a deep learning architecture based on artificial neural network (ANN). The neural network (NN) structure for crime prediction and the appropriate inputs for crime prediction is performed through: Gram-Schmidt orthogonalization (GS) for the selection of network inputs and virtual leave-one-out test (VLOO) for the selection of the optimal number of hidden neurons. Spatiotemporal distribution of the hot-spots is conducted and a methodology is developed for short-term crime forecasting using the long short-term memory (LSTM) recurrent neural networks (RNN) and convolutional neural networks (CNN).