摘要
对鳞翅目Lepidoptera和鞘翅目Coleoptera5个总科23种昆虫图像中提取昆虫面积、周长等11项数学形态特征进行了粗糙集神经网络分析,并与赵汗青统计分析加以比较,结果表明在总科阶元上,11项特征的可靠性顺序为面积、亮斑数>周长、横轴长、形状参数、圆形性、似圆度、偏心率>纵轴长、叶状性、球状性形性、似圆度、偏心率)>(纵轴长、叶状性)>(形状参数、亮斑数)。与赵汗青等人用统计学分析的结果不完全一致,但大多数属性特征重要性还是一致的。神经网络模式识别结果与传统分类结果完全一致。由此得出:粗糙集理论在昆虫依据数学形态特征进行分类方面与统计分析方法相比更为理想。
Using rough-set theory and neural network analyses of 11 math-morphological features （ MMFs ） （such as area, perimeter, etc. ） from the images of 23 species of insects of the Lepidoptera and Coleoptera fiamilles, Noctuoidea, Bombycoidea, Papilionaidea, Scarabaeoidea and Chrysomeloidea, the results are compared with those of ZHAO Han-Qing made by his statistical analysis and indicates that the ranked reliability of MMFS in the identification of insect superfamilies is： from high to low： area, hot-holenumbe 〉 perimeter, Xlength, form parameter, circttlafity, rotmdnesslikelihood, eccentricity 〉 Y-length, lobation, sphericity,roundness-likelihood, eccentricity 〉 Y-length, lobation 〉 form parameter, hot-holenumbe. The results are not completely identical with those of ZHAO Han-Qing made by his statistical analysis, but the most importance of characteristic are identical. The results of pattern recognition by neutral network are completely identical with those of traditional classifications. Accordingly, the condusion is that this theory applied in insect taxonomy is more ideahs＇tic compared with statistical analysis method, and it has great significance at superfamily level when used with rough-set neutral network.
出处
《动物分类学报》
CSCD
北大核心
2007年第1期147-152,共6页
Acta Zootaxonomica Sinica
基金
南阳师范学院青年科学研究资助项目(nytc2004k01).
关键词
昆虫分类
粗糙集
神经网络
数学形态特征
Insect taxonomy, theory of Rough Set, math-morphological features （MMF）, neutral network.
作者简介
通讯作者，E-mail：duruiqing8@163.com.
