Scholarlink

In constructing a recognition system, it is important to evaluate the features to be used in advance. In this chapter, we describe the feature evaluation. First, the meaning and necessity of feature evaluation are described, and then the ratio of between-class variance to within-class variance which is known as a simple feature evaluation method, is discussed. Then, the Bayes error, which plays an important role in feature evaluation, is introduced. The Bayes error is an extremely important concept that forms the basis of pattern recognition and machine learning, and is explained in detail in Sect. 5.3, 5.4 and 5.5. It is impossible to obtain the Bayes error directly, and its estimation methods have been studied for a long time. After introducing them, we introduce experimental results of feature evaluation using concrete data.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Feature Evaluation and Bayes Error

  • 章节

  • 发表日期：

    2026年

查看全文
  • Kenichiro Ishii,
  • Naonori Ueda,
  • Eisaku Maeda,
  • Hiroshi Murase

摘要

In constructing a recognition system, it is important to evaluate the features to be used in advance. In this chapter, we describe the feature evaluation. First, the meaning and necessity of feature evaluation are described, and then the ratio of between-class variance to within-class variance which is known as a simple feature evaluation method, is discussed. Then, the Bayes error, which plays an important role in feature evaluation, is introduced. The Bayes error is an extremely important concept that forms the basis of pattern recognition and machine learning, and is explained in detail in Sect. 5.3, 5.4 and 5.5. It is impossible to obtain the Bayes error directly, and its estimation methods have been studied for a long time. After introducing them, we introduce experimental results of feature evaluation using concrete data.