Examining the influence of cell size and bandwidth size on kernel density estimation crime hotspot maps for predicting spatial patterns of crime
Examining the influence of cell size and bandwidth size on kernel density estimation crime hotspot maps for predicting spatial patterns of crime
复制标题
DOI：
--
复制
发表时间：
2014-10
期刊：
影响因子：
--
通讯作者：
S. Chainey
中科院分区：
文献类型：
--
作者：
S. Chainey
热点地图是一种流行的技术，用于帮助确定警察巡逻和其他减少犯罪举措的目标。有许多空间分析技术可用于识别热点，但近年来最流行的是核密度估计（KDE）。 KDE 很受欢迎，因为它以视觉上吸引人的方式表示犯罪的空间分布，并且因为它被认为是常用的热点地图技术中最准确的。为了生成 KDE 输出，研究人员需要输入两个主要参数的值：像元大小和带宽大小。迄今为止，很少有人研究这些参数对 KDE 热点地图输出的影响，也没有人研究这些参数设置对热点地图的中心目的（确定未来可能发生犯罪的地方）的影响。我们通过这项研究填补了这一空白，通过使用不同的小区大小和带宽值以及住宅入室盗窃和暴力袭击的犯罪数据进行了许多实验。我们表明，单元大小对预测犯罪空间模式的 KDE 犯罪热点图影响不大，但带宽大小确实有影响。最后，我们讨论了这项研究的结果如何帮助警察从业人员和研究人员更好地利用 KDE 来制定警务和预防犯罪举措。
Hotspot mapping is a popular technique used for helping target police patrols and other crime reduction initiatives. There are a number of spatial analysis techniques that can be used for identifying hotspots, but the most popular in recent years is kernel density estimation (KDE). KDE is popular because of the visually appealing way it represents the spatial distribution of crime, and because it is considered to be the most accurate of the commonly used hotspot mapping techniques. To produce KDE outputs, the researcher is required to enter values for two main parameters: the cell size and bandwidth size. To date little research has been conducted on the influence these parameters have on KDE hotspot mapping output, and none has been conducted on the influence these parameter settings have on a hotspot map’s central purpose – to identify where crime may occur in the future. We fill this gap with this research by conducting a number of experiments using different cell size and bandwidth values with crime data on residential burglary and violent assaults. We show that cell size has little influence on KDE crime hotspot maps for predicting spatial patterns of crime, but bandwidth size does have an influence. We conclude by discussing how the findings from this research can help inform police practitioners and researchers make better use of KDE for targeting policing and crime prevention initiatives.