In this paper, time series model of ARIMA is used to make short-term forecasting of property crime for one city of China. With the given data of property crime for 50 weeks, an ARIMA model is determined and the crime amount of 1 week ahead is predicted. The model's fitting and forecasting results are compared with the SES and HES. It is shown that the ARIMA model has higher fitting and forecasting accuracy than exponential smoothing. This work will be helpful for the local police stations and municipal governments in decision making and crime suppression.

No abstract

No abstract

Though the relationships between environment and crime have been studied a lot in many countries, this work is still a void in China. This work presents the study about how property crimes are influenced by the temporal and weather factors in China. With the crime data collected from police, the property crimes pattern by season of year, day of week and time of day are investigated firstly. Then the influence of the temporal variables-major holidays, school close days and weekends-and weather on the crimes are tested. The findings show that the robbery is significantly influenced by the temporal variables but has no correlations with weather, while burglary is not only affected by the temporal variables but also correlated with sun light hours.

No abstract

No abstract

No abstract

Summary The traditional research of building fire probability analysis is from statistics or fire science. This paper combines the two methods and aims to improve the statistical method of building ignition probability determination according to the research conclusion of fire science. The specific factors that affect the ignition probability are divided into three aspects: humans, ignition sources and combustibles and environments. On this basis, the Bayesian network of building ignition probability is constructed, the nodes and conditional probability table in the Bayesian network are introduced in detail, according to which the ignition probability of building can be calculated quantitatively and objectively. Then some typical buildings are chosen as examples for the application of the method, the posterior probability value is calculated by obtaining the relevant building information and substituting them into the Bayesian network. The ignition probability is dynamic, and the comparison with the statistical data of building fire also proves its rationality.