目录
摘要： 3
Abstract: 3
第1章 导论 4
1.1选题背景 4
1.1.1数据挖掘 4
1.1.2基于生理信号的情感分析 5
1.2 研究内容及目的 6
1.2.1 研究内容 7
1.2.2 研究目的 7
第2章 数据去噪 9
2.1 数据去噪方法 9
2.2 小波去噪方法 9
第3章 基于K近邻的悲伤情感算法 12
3.1 K近邻算法对分类的影响 12
3.2 欧式距离公式 13
3.3 波形图求极大值和极小值 13
3.4 K近邻算法比对 13
第4章 结论 15
参考文献: 16
致谢： 17

基于 K近邻的悲伤情感算法的研究

摘要：
随着科技和社会的发展，数据在人类生活中也越显重要，自动的数据收集工具和成熟的数据库技术导致大量数据存放在数据库中，我们正被大量的数据淹没。如何从这些大量的数据中提取出一些对于我们有意义的数据是至关重要的，因此有了数据挖掘技术的出现。数据挖掘就是从大型数据中提取出非平凡的、有蕴涵的、先前未知的并且是潜在有用的信息，以满足人们不同应用的需要。K近邻算法（KNN）是基于统计的分类算法，是数据挖掘中一种常用的分类算法。该算法具有简单、直观、易学习等特点，备受数据挖掘研究 者们的喜爱。
本文主要研究K近邻悲伤情感算法，就是利用K近邻分类算法对人们在悲伤和平静状态下所测得的人体表的电流数据进行分类，以此来判断一个人是悲伤的，或是平静的。首先介绍了数据挖掘中的各种分类算法，详细的描述了K近邻分类算法的基本原理及其应用领域，其次介绍了数据去噪。
  针对测得数据的庞大及其K近邻算法计算量大缺陷，提出了将K近邻分类算法与数据去噪集合起来。
关键词：数据挖掘；K近邻；悲伤情感算法 ；数据去噪
Research on K nearest neighbor algorithm emotional Grief

Abstract:
With the development of science and technology and society, data is becoming more important in human life, automatic data collection tools and mature database technology lead to a lot of data stored in the database, we are being inundated with data. How to extract from these large amounts of data in some of our meaningful data it is crucial, and therefore have a data mining technologies emerge. Data mining is to extract data from a large non-trivial, there is the implication, previously unknown and potentially useful information, in order to meet the needs of people of different applications. K nearest neighbor (KNN) is based on the statistical classification algorithms, data mining is a commonly used classification algorithm. The algorithm is simple, intuitive, easy to learn and so, much of the data mining researchers alike.
  This paper studies the emotional grief K nearest neighbor algorithm is the use of K-nearest neighbor classification algorithm for people in grief and calm state of the measured current data table to classify the human body, in order to determine whether a person is sad or calm. First introduced in a variety of data mining classification algorithm, described in detail the basic principles and applications of K-nearest neighbor classification algorithm, followed by the introduction of data de-noising.
  For large data measured and K nearest neighbor algorithm large defects, proposed a K nearest neighbor classification algorithms and data de-noising set up.
Key words: Data mining; K nearest neighbor; sad emotion algorithm; data de-noising