作者：曾赟
作者介绍：广州大学法学院教授
出版日期：2018年2月9日
内容提要 当某一故意杀人犯罪发生时,侦查机关的第一要务是准确判定案件属性。其中确定案件是否属于系列杀人犯罪可谓重中之重,因为这既关涉到案件能否得以及时、准确侦破,又关乎司法公正。为便于侦查机关准确判别某一故意杀人是否属于系列杀人犯罪,兹在明确界定系列杀人犯罪概念基础上,通过聚类分析、方差分析、二元逻辑回归分析等方法创建系列杀人犯罪分类模型和回归模型。依据系列杀人犯罪分类模型,侦查机关可初步基于犯罪地是否偏僻、被害人尸体是否在犯罪现场、凶犯除故意杀人是否还有抢劫、是否有犯罪前科等四个方面直观判定某一故意杀人是否为系列杀人犯罪;依据系列杀人犯罪回归模型,侦查机关可精准计算出某一故意杀人为系列杀人犯罪的概率值。分类模型的直观性和回归模型的精准性为侦查机关判别故意杀人犯罪案件属性提供了一个可具体测量的科学依据。
关键词 系列杀人 数据编码 聚类分析 方差分析 二元逻辑回归分析
Abstract: When an intentional homicide crime happens, the first task of investigation organs is to judge whether it is a serial murder, which is the most important, because it is not only related to whether investigation organs can solve the murder case correctly and promptly, but also related to judicial justice. In order to judge scientifically whether the murder is a serial case, on the basis of clearing the definition of serial murder, the classification model and regression model of the serial murder will be created through the cluster analysis, analysis of variance and dualistic logistic regression analysis. According to the classification model of serial murder, investigation organs can simply and directly judge whether the murder is a serial one from whether the crime place is isolated or not, whether the corpse is at the crime scene or not, whether there is a robbery or not, whether criminals have the criminal history or not. According to the regression model of serial murder, investigation organs can accurately calculate the probability of whether the murder is a serial one. The intuition of classification model and the accuracy of regression model provide a measurable and scientific foundation for investigation organs to judge the attribute of serial murder.