首页> 外国专利> LOGISTIC REGRESSION ANALYSIS SYSTEM AND LOGISTIC REGRESSION ANALYSIS PROGRAM

LOGISTIC REGRESSION ANALYSIS SYSTEM AND LOGISTIC REGRESSION ANALYSIS PROGRAM

机译:Logistic回归分析系统和Logistic回归分析程序

摘要

PROBLEM TO BE SOLVED: To provide a logistic regression analysis system and a logistic regression analysis program for performing logistic regression analysis by using the variable category of explanation variables corresponding to a data category whose composition ratio in a data item is small.;SOLUTION: An abnormality detection system 10 as a logistic regression analysis system includes: a determination part 15a for determining whether or not it is possible to input the variable category of explanation variables to a logistic regression function on the basis of the composition ratio of the data category of a data item; a probability calculation part 15b for calculating probability that an event has occurred in the past by using the data category of the data item corresponding to the variable category whose input is determined to be impossible as an object; an integer integration part 15c for integrating this probability into an integer by a predetermined multiple; and a development part 15d for generating the conversion value of the variable category of the explanation variables by developing the probability integrated into an integer on the basis of the validity of the explanation variables of the variable category.;COPYRIGHT: (C)2012,JPO&INPIT
机译:解决的问题:提供一种逻辑回归分析系统和逻辑回归分析程序,该逻辑回归分析系统和逻辑回归分析程序通过使用与数据项的组成比较小的数据类别相对应的解释变量的变量类别来执行逻辑回归分析。作为逻辑回归分析系统的异常检测系统10包括:确定部分15a,用于基于a的数据类别的组成比来确定是否可以将解释变量的变量类别输入到逻辑回归函数。数据项;概率计算部分15b,用于通过使用与确定为不可能输入的变量类别相对应的数据项的数据类别作为对象来计算过去发生事件的概率。整数积分部分15c,用于将该概率积分为预定倍数的整数;展开部15d,其根据变量类别的解释变量的有效性,通过将被积分为整数的概率进行展开,来生成解释变量的变量类别的转换值。COPYRIGHT:(C)2012,JPO&INPIT

著录项

  • 公开/公告号JP2012173899A

    专利类型

  • 公开/公告日2012-09-10

    原文格式PDF

  • 申请/专利权人 SCSK CORP;

    申请/专利号JP20110034012

  • 发明设计人 OTA KAORU;NOMA MIHOKO;

    申请日2011-02-18

  • 分类号G06Q20/24;G06Q90/00;

  • 国家 JP

  • 入库时间 2022-08-21 17:42:41

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号