New models and efficient algorithms for hazard detection

Abstract

Many environmental factors such as deficiency of some elements (certain vitamins), radioactive contamination accidents, infectious disease epidemics and so on can make people sick. Finding out the possible positions of environmental hazards is very important because it can help researchers to identify the causes of environmental hazards and furthermore to remove them. However, it is not an easy task without the help of computers because the amount of information people must handle and analyze is massive in modern society. In order to help to find out hazards by computers, many models and algorithms are designed. For example, analysis of clusters of diseases is an important method. However, current cluster analysis methods ignore mobility of people, which is an important feature of modern society. Therefore, the methods cannot pinpoint the exact areas responsible for the development of a disease and how much possibilities hazards appear in the areas. In this paper, we propose novel models and algorithms based on the patients' residential history, mobility of locations of people and disease principles. Our models and algorithms differ from previous ones in the following ways. First, we consider more complex situations such as multiple hazards and outliers and so on. Second, our algorithms not only output the possible areas responsible for diseases but also output the how much possibilities hazards appear in the areas.