A spatio-temporal clustering method for irregular shapes based on maximum correlation and risk deviation

Abstract

The invention relates to a spatio-temporal scanning clustering method for irregular shapes based on maximum correlation and risk deviation, belonging to the field of spatio-temporal event clustering analysis method. The method comprises the steps of firstly obtaining the graphic structure of the research area according to the case onset location and traffic route in the research area; then, according to the maximum correlation (mlink) algorithm,determining the scan window shape and searching the potential aggregation region; then calculating the relative risk and risk deviation of the aggregation to determine whether the relative risk of the aggregation is accurate or not. Compared with the prior art, the method of the invention changes the scanning window of the conventional cylindrical spatio-temporal scanning method, and the spatial base thereof can be irregular in shape, thereby facilitating the discovery of the spatial aggregation of unconventional shapes. And the relative risk bias assessment is introduced to assess the risk of aggregation. If the deviation is too large, the region can be further observed to avoid causing unnecessary loss of early warning and increase the reliability of disease prevention and early warning in the region.

Description

technical field [0001] The invention relates to a space-time scanning clustering method for irregular shapes based on maximum correlation and risk deviation, and belongs to the field of space-time event clustering analysis methods. Background technique [0002] In disease detection and prevention, spatio-temporal scanning clustering methods are often used for early exploration of disease outbreaks, and many scientists use spatio-temporal scanning clustering methods to explore and evaluate geographic clustering of diseases. However, in the spatio-temporal scanning clustering method, the scanning window for defining disease clusters is often cylindrical, and its scanning window is often not dynamically changed with time. The scanning clustering method cannot accurately and timely detect the aggregation of the disease; and in previous studies, the relative risk assessment of the detected aggregation was not considered, although the accuracy of the spatio-temporal scanning model...

AI Technical Summary

Problems solved by technology

However, in the spatio-temporal scanning clustering method, the scanning window for defining disease clusters is often cylindrical, and its scanning window is often not dynamically changed with time. The scanning clustering method cannot accurately and timely detect the aggregation of the disease; and in previous studies, the relative risk assessment of the detected aggregation was not considered, although the accuracy of the spatio-temporal scanning model to detect the spatio-temporal aggregation Higher, but there is still the possibility of errors. If the wrong aggregation phenomenon is detected, it will cause unnecessary loss of disease early warning, cause unnecessary panic among the people in the area, and reduce the credibility of disease early warning.

Images