Learning distributed representations of concepts from relational data using linear relation.pdfVIP

下载本文档

10
0
约5.33万字
约 22页
2017-04-11 发布于江苏
举报
版权申诉

Learning distributed representations of concepts from relational data using linear relation.pdf

1、原创力文档（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。。
2、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
3、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。
4、该文档为VIP文档，如果想要下载，成为VIP会员后，下载免费。
5、成为VIP后，下载本文档将扣除1次下载权益。下载后，不支持退款、换文档。如有疑问请联系我们。
6、成为VIP后，您将拥有八大权益，权益包括：VIP文档下载权益、阅读免打扰、文档格式转换、高级专利检索、专属身份标志、高级客服、多端互通、版权登记。
7、VIP文档为合作方或网友上传，每下载1次，网站将根据用户上传文档的质量评分、类型等，对文档贡献者给予高额补贴、流量扶持。如果你也想贡献VIP文档。上传文档

Learning distributed representations of concepts from relational data using linear relation

Gatsby Computational Neuroscience Unit 17 Queen Square, London University College London WC1N 3AR, United Kingdom http://www.gatsby.ucl.ac.uk +44 20 7679 1176 Funded in part by the Gatsby Charitable Foundation. March 18, 2000 GCNU TR 2000–002 Learning Distributed Representations of Concepts using Linear Relational Embedding Alberto Paccanaro Geoffrey Hinton Gatsby Unit Abstract In this paper we introduce Linear Relational Embedding as a means of learn- ing a distributed representation of concepts from data consisting of binary relations between concepts. The key idea is to represent concepts as vectors, binary relations as matrices, and the operation of applying a relation to a concept as a matrix-vector multiplication that produces an approximation to the related concept. A representation for concepts and relations is learned by maximizing an appropriate discriminative goodness function using gradient ascent. On a task involving family relationships, learning is fast and leads to good generalization. Learning Distributed Representations of Concepts using Linear Relational Embedding Alberto Paccanaro Geoffrey Hinton Gatsby Unit 1 Introduction Given data which consists of concepts and relations among concepts, our goal is to correctly pre- dict unobserved instances of relationships between concepts. We do this by representing each concept as a vector in a Euclidean space and the relationships between concepts as linear opera- tions. To illustrate the approach, we start with a very simple task which we call the number prob- lem. The data consists of integers and operations among integers. In the modular number prob- lem the numbers are integers in the set V = [0 : : : m 1] and the set of operations is R = f+1;1;+2;2;+3;3;+4;4;+0gm , where the subscript indicates that the operations are per- formed modulo m. The data then consists of all or some of the triplets (num1; op; num2) where num1; num2 2 V , op 2 R, and num2 is the result of applying operation op to numb