大数据: 预知未来的高科技水晶球

栏目:业绩展示

更新时间:2021-02-01

浏览: 72979

大数据: 预知未来的高科技水晶球

产品简介

Early last year,you might recall, targetfounditselfatthecenterofastormofoutrage.theretailersnumbercrunchershadcomeupwithastatisticalmethodforpredictingwhich itss obecomepregnantinthenearfuture,givingtargetsmarketersaheadstartonpitchingthembabyproducts .你这是这家零售商的数据专家们,哪个顾客近themodelworked : targetexpandeditscustomerbaseforpregnancyandinfant-careproductsbyabout 30 %.butthemediabrouhaha, witheveryonefromthenewyorktimestofoxnewsaccusingthecompanyofspyingonshoppers,took weeks to die down .该模型是在塔吉特销售孕期和婴幼儿产品的顾客但这引起了舆论,从《纽约时报》 (The New York Times )到foxnews(foxnews ),所有人都指责他是“暗中监视”该公司的购物者。

产品介绍

本文摘要:Early last year,you might recall, targetfounditselfatthecenterofastormofoutrage.theretailersnumbercrunchershadcomeupwithastatisticalmethodforpredictingwhich itss obecomepregnantinthenearfuture,givingtargetsmarketersaheadstartonpitchingthembabyproducts .你这是这家零售商的数据专家们,哪个顾客近themodelworked : targetexpandeditscustomerbaseforpregnancyandinfant-careproductsbyabout 30 %.butthemediabrouhaha, witheveryonefromthenewyorktimestofoxnewsaccusingthecompanyofspyingonshoppers,took weeks to die down .该模型是在塔吉特销售孕期和婴幼儿产品的顾客但这引起了舆论,从《纽约时报》 (The New York Times )到foxnews(foxnews ),所有人都指责他是“暗中监视”该公司的购物者。

Early last year,you might recall, targetfounditselfatthecenterofastormofoutrage.theretailersnumbercrunchershadcomeupwithastatisticalmethodforpredictingwhich itss obecomepregnantinthenearfuture,givingtargetsmarketersaheadstartonpitchingthembabyproducts .你这是这家零售商的数据专家们,哪个顾客近themodelworked : targetexpandeditscustomerbaseforpregnancyandinfant-careproductsbyabout 30 %.butthemediabrouhaha, witheveryonefromthenewyorktimestofoxnewsaccusingthecompanyofspyingonshoppers,took weeks to die down .该模型是在塔吉特销售孕期和婴幼儿产品的顾客但这引起了舆论,从《纽约时报》 (The New York Times )到foxnews(foxnews ),所有人都指责他是“暗中监视”该公司的购物者。这场风波几周后才平息。iftargetssuccessatsettingitssightsonpotentialmoms-to-begivesyouthecreeps, ericsiegelsnewbookcouldruinyourwholeday.siegelisaformercolumbiaprofessorwhosecompany,Predictive Impact, buildsmathematicalmodelsthatcullvaluablenuggetsofdatafromfloodsofrawinformation.companiesusethetoolstoforecasteverythingfrom w to which movies well watch,tohowlikelywearetobeinacaraccidentordefaultonourcreditcards .如果塔吉特很好地监视准妈妈,早就让你毛骨悚然了。西格尔的新书不会让你不安吧? 西格尔曾经是哥伦比亚大学(Columbia University )的教授,该公司被称为“影响预测”(Predictive Impact ),专门开发了能够从大量原始数据中提取非常有价值的信息的数学模型。

各类公司正在使用这些工具展开预测。我们想卖什么,我们想看什么样的电影,不管我们遭遇交通事故的可能性有多低,我们都可以预测出我们有可能欠信用卡的债。

亚博App下载

inpredictiveanalytics 3360 thepowertopredictwhowillclick,Buy,Lie,or Die,siegelexplainshowthesemodelsworkandwhere in clear,coller predictive analytics,or PA, isthescienceoflearningfromexperience.startingwithdataaboutthepastandcurrentbehaviorofagivengroupofpeople-- whethercussion Patie prison inmates up for parole,voters,or employees-- analystscanpredictwhattheyllprobablydonext .是《预测分析:预测谁将页面、出售、骗子或丧生的力量》 (predictive analytics 3360 根据既定人群3354的顾客、患者、减刑囚犯、选民或员工——的过去和现在的不道德数据,分析师可以预见到他们下一步可能会变得不道德。thiskindofhigh-techcrystalballisbehindthegrowingtrendtomakedecisionsmoredatadriven, Siegel writes. In fact是anorganizationthatdoesntleverageitsdatainthiswayislikeapersonwithaphotographicmemorywhoneverbotherstothi 西格尔写道,“每天流行,依靠数据进行决策的倾向”的背后。“实际上,如果某个组织不用这种方法最大限度地利用自己的数据,就不会因为一个人有浏览的能力而烦恼”。

predictiveanalyticsispackedwithexamplesofhowciti,Facebook,Ford,IBM,Google, Netflix是paypalandmanyotherbusinessesandgovernmentagencieshaveputpatowork.PFizer,for instance, hasapredictivemodeltoforetellthelikelihoodthatapatientwillrespondtoagivennewdrugwithinthreeweeks.linkedinusespatopinpointto fee sconnections.attheirs表示amathematicalrankingsystemappliedtopasttaxreturnsempoweredirsanalyststofind 25 timesmoretaxevasion, withoutincreasingthenumberofinvestigations .这本书展示了非常丰富的案例,涉及花旗集团(Citi )、Facebook、IBM、Google(Google )、网飞。例如,辉瑞制药(Pfizer )具有预测模型,预告患者在3周内对等效新药发生药效反应的概率。LinkedIn会使用PA准确寻找想要联系的用户。在美国国税局(IRS ),过去作为纳税申报书使用的数学排名系统“使IRS的分析师可以找到以前的25倍以上的洗钱”。

andthenthereshewlett-packard.acoupleofyearsago、alarmedbyannualturnoverratesinsomedivisionsashighas 20 %、hpdecidedtotryanticic 000 employeesworldwideweremostlikelytoquit.beginningwithreamsofdataonthingslikesalaries,raises,promotions, andjooins ateamofanalystscorrelatedthatinformationwithdetailedemploymentrecordsofpeoplewhohadalreadyleft.basedonthesimilarities ignedeachcurrentemployeeaflightriskscore .也有Hewlett-Packard的案例。几年前,惠普的一些部门每年离职率达到20%,但惠普要求预测全球33万名员工中谁最有可能打招呼。分析师小组需要根据报酬水平、加薪情况、晋升情况、轮调情况等大量数据,结合退休员工的详细经验积极进行分析。根据他们找到的数据的相似性,研究者们给现在的所有员工评定了退休风险得分。


本文关键词:亚博APP,,亚博App下载,亚博手机版

本文来源:亚博APP-www.fct518.com