A Comparative Evaluation of Spatio Temporal Deep Learning Techniques for Crime Prediction

EasyChair Preprint 5648

Abstract

This paper presents a detailed evaluation of three
spatiotemporal deep learning architectures for crime prediction.
These network architectures are as follows: the Spatio Temporal
Residual Network (ST-ResNet), the Deep Multi-View Spatio
Temporal Network (DMVST-Net), and the Spatio Temporal Dynamic
Network (STD-Net). The architectures were trained using
Chicago crime data set. The Root Mean Square Error (RMSE)
and Mean Absolute Error (MAE) were used as performance
metrics to evaluate the model. Results show that the STD-Net
achieved the best results with an RMSE of 0.2870, and MAE
of 0.2093, while the DMVST-Net achieved an RMSE of 0.4171
and an MAE of 0.3455. The ST-ResNet achieved and an RMSE
of 0.4033 and an MAE of 0.3278. Future work will include
training these algorithms with crime data augmented
with external data such as climate and socioeconomic data. We also
will explore hyperparameter optimization of these algorithms
using techniques such as evolutionary computation.

Keyphrases: DMVST-Net, ST-ResNet, STD-Net, crime prediction, spatio-temporal

@booklet{EasyChair:5648,
  author    = {Tawanda Matereke and Clement Nyirenda and Mehrdad Ghaziasgar},
  title     = {A Comparative Evaluation of Spatio Temporal Deep Learning Techniques for Crime Prediction},
  howpublished = {EasyChair Preprint 5648},
  year      = {EasyChair, 2021}}