Teaching AI about human knowledge

Teaching AI about human knowledge

ID:41466532

大小:584.75 KB

页数:19页

时间:2019-08-25

Teaching AI about human knowledge_第1页
Teaching AI about human knowledge_第2页
Teaching AI about human knowledge_第3页
Teaching AI about human knowledge_第4页
Teaching AI about human knowledge_第5页
资源描述:

《Teaching AI about human knowledge》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、TeachingAIabouthumanknowledgeInesMontaniExplosionAIExplosionAIisadigitalstudiospecialisinginArtificialIntelligenceandNaturalLanguageProcessing.Open-sourcelibraryforindustrial-strengthNaturalLanguageProcessingspaCy’snext-generationMachineLearninglibraryfordeeplearningwithtextcomingso

2、on:pre-trained,customisablemodelsDataStoreforavarietyoflanguagesanddomainsMachinelearningisprogrammingbyexample.Examplesareyoursourcecode,trainingiscompilation.exampleslabelsinputpredictiontrainingdrawexamplesfromthesamedistributionastheruntimeinputsgoal:system’spredictiongivensome

3、inputmatcheslabelahumanwouldhaveassignedHowmachines“learn”Example:trainingasimplepart-of-speechtaggerwiththeperceptronalgorithm(teachthemodeltorecogniseverbs,nouns,etc.)deftrain_tagger(examples):examples=words,tags,contextsW=defaultdict(lambda:zeros(n_tags))theweightswe'lltrainfor(

4、word,prev,next),human_taginexamples:scores=W[word]+W[prev]+W[next]scoreeachtaggivenweights&contextguess=scores.argmax()getthebest-scoringtagifguess!=human_tag:iftheguesswaswrong,adjustweightsforfeatin(word,prev,next):W[feat][guess]-=1decreasescoreforbadtaginthiscontextW[feat][human

5、_tag]+=1increasescoreforgoodtaginthiscontextThebottleneckinAIisdata,notalgorithms.Algorithmsaregeneral,trainingdataisspecific.dataquality,dataquantityandaccuracyproblemsarestillthebiggestproblemsinAI(Source:TheStateofAIsurvey)youcanextractknowledgefromallkindsofsources,e.g.sentiment

6、fromemojionReddit�youusuallyneedatleastsomedataspecifictoyourproblemannotatedbyhumansWherehumanknowledgeinAIreallycomesfromMechanicalTurkhumanannotators~$5perhourboringtaskslowincentivesImages:AmazonMechanicalTurk,depressing.orgDon’texpectgreatdataifyou’reboringtheshitoutofunderpaid

7、people.Whyarewe“designingaround”this?“TakingaHIT:DesigningaroundRejection,Mistrust,Risk,andWorkers’ExperiencesinAmazonMechanicalTurk”(McInnisetal.,2016)datacollectionneedsthesametreatmentasallotherhuman-facingprocessesgoodUX+purpose+incentives=betterqualitySOLUTION#1UX-drivendataco

8、llectionwithactivelearning

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。