模拟退火组合优化法在模式识别中的若干应用
Some Applications of Simulated Annealing to Pattern Recognition
-
摘要: 本文将用于求解组合优化问题的模拟退火法引入聚类分析、属性关系图同态、分段曲线拟 合和特征选择等模式识别问题.(1)提出了一类新的聚类分析算法--模拟退火聚类法; (2)给出了一种模拟退火图同态的方案和实现算法--ALISOM;(3)详细地讨论了如何应 用模拟退火组合优化法进行分段曲线拟合和特征选择.Abstract: Simulated annealing technique for solving combinatorial optimization problems has been applied to cluster analysis, isomorphisms of attributed relational graphs, piecewise curve fitting and feature selection.(1).A class of new clustering algorithms by simulated annealing are presented.(2). The problem of isomorphisms of attributed relational graph is treated by annealing simulation. An annealing isomorphism algorithm ALISOM is presented. (3).The applications of simulated annealing technique to piecewise curve fitting and feature selection are discussed in detail.
-
Key words:
- Simulated annealing /
- pattern recognition /
- cluster analysis /
- graph isomorphism /
- feature selection
计量
- 文章访问数: 2194
- HTML全文浏览量: 145
- PDF下载量: 1062
- 被引次数: 0