1. 1、本文档共24页,可阅读全部内容。
  2. 2、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。
  3. 3、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
  4. 4、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)。
查看更多
NN13

BUS3650 / BUS4650 / BUS5650 Business Applications of Neural Networks Lecture 13 ó Dr. Kate A. Smith, 1999 This Lecture Video Future Directions Revision Exam Announcement Reminder: Assignments due this Friday school office by 12 noon 3rd years: hand in disk containing your answers and all NShell2 files from your best result, plus 3 pages of experiments postgrads: hand in a disk containing all NShell2 files created from your best NN result, plus your 4-6 page paper We must be able to reproduce your results ... Teaching and Subject Evaluation Forms Please take a few minutes to provide some feedback to me about the course my teaching of it Any positive and (constructive) negative feedback is welcome, and will be considered when I next teach the course Video Presentation “Learning Systems” by Professor Bernard Widrow (Stanford University) 20 minutes from a 2 hour video interview with Widrow in 1960’s program “Science in Action” Video available from Caulfield campus library: 621.38 LEA The Future of Neural Networks In the first module, we looked at some of the application areas and companies that are using neural networks and data mining Clearly, neural networks have very strong support (and high expectations) from business & industry We must be careful about such expectations it is easy to get carried away with the idea of an artifical brain replacing human activities neural networks cannot do everything Current Research learning & training: trying to get better performance & faster autonomous learning: systems which are unsupervised & learn “on-line” in real time, which can be used to replace humans (in dangerous situations for example) biological models: much research into how the brain works which architecture for which problem? Emerging Directions Speed & Storage need advances in hardware design this will be market-driven (smaller & faster) already we have neural chips for MFNN & self-organising neural networks (& others) also Field Programmable Gate Arrays (FPGA

文档评论(0)

saa9099 + 关注
实名认证
内容提供者

该用户很懒,什么也没介绍

1亿VIP精品文档

相关文档