• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Artificial Intelligence and Pattern Recognition >Adaptive and Efficient Classifier for Data Streams
【24h】

Adaptive and Efficient Classifier for Data Streams

机译：用于数据流的自适应和有效分类器

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

As data streams are gaining prominence in a growing number of emerging application domains, classification of data streams is becoming an active research area. The major challenges in data stream mining are (i) infinite volume of stream data and its fast speed as usually stream classification process can only pass data stream once, (ii) to incorporate evolving features of data streams. Classification of data streams require algorithms that are efficient, fast, use less memory and easily adaptable to concept drift. To overcome these problems, this paper proposes and implements an ensemble boosting approach to develop high speed classification algorithm on streaming data with concept drifts. This ensemble approach improves both the efficiency in model learning, accuracy in performing classification and adapts concept drift, to result a better classifier. The technique is based on adaptive sliding window and adaptive size hoeffding tree.
机译：随着数据流在越来越多的新兴应用领域中获得突出，数据流的分类正在成为积极的研究区域。数据流挖掘中的主要挑战是（i）无限量的流数据及其快速速度，因为通常流分类过程只能通过数据流一次，（ii）结合数据流的不断变化的特征。数据流的分类需要高效，快速，使用较少的内存并且容易适应概念漂移的算法。为了克服这些问题，本文提出并实现了一个集合升压方法，以在具有概念漂移的流数据上开发高速分类算法。该集合方法提高了模型学习的效率，执行分类的准确性和适应概念漂移，以产生更好的分类器。该技术基于自适应滑动窗口和自适应尺寸的Hoeffd树。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号