Maximum Entropy Analysis for Pattern Recognition
Maximum Entropy Analysis for Pattern Recognition
复制标题
模式识别的最大熵分析
DOI：
10.1007/978-94-009-0683-9_27
复制
发表时间：
1990
期刊：
影响因子：
--
通讯作者：
C. Chen
中科院分区：
文献类型：
--
作者：
C. Chen
模式识别中的特征提取、分类、聚类和学习都与最大或最小熵原理密切相关。这种关系在本文中进行了审查。然后强调需要使用神经网络的自适应模式识别。神经网络和传统的统计分类器之间的比较。
Feature extraction, classification, clustering and learning in pattern recognition are closely related to the maximum or minimum entropy principles. Such relationships are reviewed in this paper. The need for adaptive pattern recognition using neural networks is then emphasized. A comparison between neural networks and conventional statistical classifiers is also presented.