摘要
提出了基于数据挖掘技术及基于支持向量机的两种水中目标识别方法 ,分别藉助目标噪声特征量提取和模式识别算法以及支持向量及二次规划算法 ,对比性地研究了不同工况下三类目标的分类识别效果。其方法和结果对水中目标识别有较好的参考价值。
This paper deals with the underwater target recognition approach based on data mining technique and support vector machine. By means of target noise characteristic abstraction, pattern recognition algorithm, support vector machine and quadratic programming algorithm, the paper completed the clustering analysis of three kinds of targets at different ambient background situation. Experiment results indicate that this method has good performance and robustness, and the recognition result is satisfactory for practical use.
出处
《计算机与数字工程》
2004年第6期41-44,72,共5页
Computer & Digital Engineering
关键词
数据挖掘
支持向量机
目标识别
聚类分析
data mining, support vector machine, target recognition, clustering analysis