摘要
特征选择是机器学习及模式识别领域的重要问题之一。针对高维数据对象,特征选择不仅可以在保证数据完整性的情况下减少特征维数,还能够提高分类精度。文中提出了一种基于自适应模拟退火遗传算法的特征选择方法,该方法将模拟退火算法嵌入到自适应遗传算法的循环体中,利用模拟退火算法具有较强的局部搜索能力,并且能够使搜索过程避免陷入局部最优解的特点,解决了基本遗传算法收敛速度慢,时间复杂度高的缺点。实验结果表明,在保证分类正确率的前提下,该方法有效提高了特征选择效率。
Feature selection is one of important problems in machine learning and pattern recognition areas.For high demensian data,feature dimension can be decreased under the condition of ensuring data integrity and classification accuracy can be improved by feature selection.A feature selection method based on adaptive simulated annealing genetic algorithm was proposed,which embeds the simulated annealing algorithm in the circle of adaptive genetic algorithm and uses the feature that simulated annealing algorithm has the strong ability of local search and makes searching process avoid sinking into the local optimal solution,to solve the shortcomings of slow convergence speed and high time complexity.The experiment results show that the method can guarantee the correct rate of classification and improve the efficiency of feature selection.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2009年第1期81-85,共5页
Acta Armamentarii
基金
国防基础科研项目(C1120060497-06-02)
关键词
人工智能
特征选择
自适应遗传算法
模拟退火算法
搜索能力
artificial intelligence
feature selection
adaptive genetic algorithm
simulated annealing algorithm
search ability