Series Crime Prediction Research Based on Regional Weighted Models

Serial criminal cases directly threaten the security of the people and even the country and society. At present, grass-roots public security work still adopts the method of "handling cases by experience", which is difficult to popularize. In view of this situation, combined with the existing scientific researches, we innovate and optimize the regional coverage weighted model on the basis of the technology of criminal geographical portrait and the high regularity of serial criminal cases. Through data fitting and geographical risk probability weighting, we have carried out geometric mapping and constructed a visual serial criminal prediction models, which effectively break through the singularity and limitations of the original crime prediction and significantly improve the prediction accuracy.