188宝金博页面版

  • 图案背景
  • 纯色背景
视图
标记
批注
批注本地保存成功，开通会员云端永久保存 去开通
adsuhviusa

上传于：2018-09-02

粉丝量：104

该文档贡献者很忙，什么也没留下。


  • 相关
  • 目录
  • 笔记
  • 书签

188宝金博页面版:更多相关文档

  • 【精品】deep与deeply

    星级： 25 页

  • Deep sea, deep snow … deep space

    星级： 20 页

  • deep

    星级： 7 页

  • СЕМАНТИЧЕСКАЯ ДЕРИВАЦИЯ ПРИЛАГАТЕЛЬНОГО DEEP

    星级： 5 页

  • 类似deep与deeply副词的区别

    星级： 1 页

  • DEEP JEEP

    星级： 4 页

  • 進深DEEPER, DEEPER

    星级： 8 页

  • Deep sea in deep trouble

    星级： 3 页

  • 基于Deep

    星级： 3 页

  • a deep

    星级： 7 页

暂无目录

点击鼠标右键菜单,创建目录

    暂无笔记

    选择文本，点击鼠标右键菜单，添加笔记

    暂无书签

    在左侧文档中，点击鼠标右键，添加书签

    188宝金博页面版: Deep Neural Networks

    下载积分： 500

    内容提示： Deep Neural NetworksRandall Balestriero and Richard G. BaraniukRice UniversityDecember 2, 2017AbstractDeep Neural Networks (DNNs) are universal function approximators providing state-of- the-artsolutions on wide range of applications. Common perceptual tasks such as speech recognition, im-age classif ication, and object tracking are now commonly tackled via DNNs. Some fundamentalproblems remain: (1) the lack of a mathematical framework providing an explicit and interpretableinput-output formula for any top...

    文档格式：PDF | 页数：75 | 浏览次数：31 | 上传日期：2018-09-02 02:06:24 | 文档星级：
    Deep Neural NetworksRandall Balestriero and Richard G. BaraniukRice UniversityDecember 2, 2017AbstractDeep Neural Networks (DNNs) are universal function approximators providing state-of- the-artsolutions on wide range of applications. Common perceptual tasks such as speech recognition, im-age classif ication, and object tracking are now commonly tackled via DNNs. Some fundamentalproblems remain: (1) the lack of a mathematical framework providing an explicit and interpretableinput-output formula for any topology, (2) quantif ication of DNNs stability regarding adversarial ex-amples (i.e. modif ied inputs fooling DNN predictions whilst undetectable to humans), (3) absence ofgeneralization guarantees and controllable behaviors for ambiguous patterns, (4) leverage unlabeleddata to apply DNNs to domains where expert labeling is scarce as in the medical f ield. Answeringthose points would provide theoretical perspectives for further developments based on a commonground. Furthermore, DNNs are now deployed in tremendous societal applications, pushing theneed to f ill this theoretical gap to ensure control, reliability, and interpretability.1arXiv:1710.09302v3 [stat.ML] 6 Nov 2017

    阅读了该文档的用户还阅读了这些文档

    188宝金博页面版:关于我们

    188宝金博页面版:帮助中心

    188宝金博页面版:关注我们

    • 新浪微博

    关注188宝金博页面版公众号

    188宝金博页面版
    阅读
    APP
    阅读
    返回
    顶部
    188宝金博页面版官网登录在线平台入口（2026已更新）—江苏协昌电子科技股份有限公司