DISTRIBUTED CLASSIFIER CHAIN OPTIMIZATION FOR REAL-TIME MULTIMEDIA STREAM MINING SYSTEMS.pdf

DISTRIBUTED CLASSIFIER CHAIN OPTIMIZATION FOR REAL-TIME MULTIMEDIA STREAM MINING SYSTEMS.pdf

  1. 1、本文档共9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
DISTRIBUTED CLASSIFIER CHAIN OPTIMIZATION FOR REAL-TIME MULTIMEDIA STREAM MINING SYSTEMS

DISTRIBUTED CLASSIFIER CHAIN OPTIMIZATION FOR REAL-TIME MULTIMEDIA STREAM MINING SYSTEMS Brian Foo, Mihaela van der Schaar University of California Los Angeles (UCLA), Dept. of Electrical Engineering (EE), Los Angeles, CA, 90095 Tel: +1-713-818-9047, Fax: +1-310-206-4685, Email: {bkungfoo , mihaela}@ ABSTRACT We consider the problem of optimally configuring classifier chains for real-time multimedia stream mining systems. Jointly maximizing the performance over several classifiers under minimal end-to-end processing delay is a difficult task due to the distributed nature of analytics (e.g. utilized models or stored data sets), where changing the filtering process at a single classifier can have an unpredictable effect on both the feature values of data arriving at classifiers further downstream, as well as the end-to-end processing delay. While the utility function can not be accurately modeled, in this paper we propose a randomized distributed algorithm that guarantees almost sure convergence to the optimal solution. We also provide results using speech data showing that the algorithm can perform well under highly dynamic environments. Keywords: Distributed systems, Queuing theory, Multimedia stream mining, Multiagent systems 1. INTRODUCTION Recently, data mining applications have emerged that require operations such as classification, filtering, aggregation, and correlation over high-volume, continuous multimedia streams in real time [1]. Due to the naturally distributed set of data sources and jobs, as well as high computational burdens for the analytics, distributed mining systems have been recently developed [2]. These systems leverage computational resources from a set of distributed processing nodes and provide the framework to deploy and run stream mining applications on various resource topologies [2]. A key challenge in distributed, real-time multimedia stream mining systems is how to cope effectively with system overl

文档评论(0)

l215322 + 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