摘要
针对神经元的空间几何形态特征分类问题以及神经元的生长预测问题进行了探讨.结合神经元的形态数据,分别建立了基于支持向量机的神经元形态分类模型、基于主成分分析和支持向量机的神经元分类模型以及基于遗传算法和RBF网络的神经元生长预测模型,在较合理的假设下,对各个模型进行求解,得到了较理想的结果.
This paper studies the classification of spatial geometry morphology about neuronal and the prediction of the growth of the neuronal. Based on the morphological data of neurons, we established a model for morphological classification of neurons based on support vector machine, and a model for morphological classification of neurons based on principal component analysis and support vector machine, based on the RBF network and genetic algorithms, the neuronal growth prediction model also be build up. In the reasonable assumption, each model is solved and we get the better results.
出处
《数学的实践与认识》
CSCD
北大核心
2011年第14期149-155,共7页
Mathematics in Practice and Theory
关键词
神经元
几何形态特征
分类
支持向量机
主成分分析
neurons
Geometry characteristics
categorization
support vector machine
principal component analysis
