• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>International Journal of Innovative Computing Information and Control >EVALUATION OF DISCRIMINATION POWER OF FEATURES IN THE PATTERN CLASSIFICATION PROBLEM USING ARIF INDEX AND ITS APPLICATION TO PHYSIOLOGICAL DATASETS
【24h】

EVALUATION OF DISCRIMINATION POWER OF FEATURES IN THE PATTERN CLASSIFICATION PROBLEM USING ARIF INDEX AND ITS APPLICATION TO PHYSIOLOGICAL DATASETS

机译：基于ARIF指数的图案分类问题中特征识别力的评估及其在生理数据中的应用

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In this paper, a novel index called Arif Index is proposed to evaluate the discrimination power of the features in pattern classification. Optimizing the performance, of a classifier requires a prior knowledge of maximum achievable accuracy in the pattern classification using a particular set of features. Moreover, it is also desirable to know that this set of features is separable by a decision boundary of any arbitrary complexity or not. Proposed index is model free and requires no clustering algorithm to discover the clustering structure present in the feature space. It is only based on the information of local neighborhood of feature vectors in the feature space. This index can be used to predict the classification accuracy and density of the feature vectors of a class in the feature space. It was found in this paper that predicted accuracy and Arif index are very strongly correlated with each other (R~2 = 0.99 with p-value nearly equals to zero). This index is designed to predict the maximum achievable accuracy by a particular set of features. Implementation of the index is simple and time efficient. Performance of Arif index on different benchmark physiological data sets is found to be in consistent with the reported accuracies in the literature. Hence, this index will be very useful in providing prior useful information about the quality of features before designing any classifier.
机译：本文提出了一种新的索引，称为Arif索引，以评估特征在模式分类中的辨别力。要优化分类器的性能，需要使用特定的一组功能在模式分类中获得最大可实现精度的先验知识。此外，还希望知道该组特征可以被任意任意复杂度的判定边界所分离。拟议的索引是无模型的，不需要聚类算法即可发现特征空间中存在的聚类结构。它仅基于特征空间中特征向量的局部邻域信息。该索引可用于预测特征空间中一类特征向量的分类准确性和密度。在本文中发现，预测的准确性和Arif指数之间具有非常强的相关性（R〜2 = 0.99，p值几乎等于零）。该指数旨在通过一组特定功能预测最大可达到的精度。索引的实现既简单又省时。发现在不同基准生理数据集上的Arif指数表现与文献中报道的准确性一致。因此，该索引对于在设计任何分类器之前提供有关特征质量的先前有用信息非常有用。

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