Then it presented cover based constructive neural networks,and compared it with NB and KNN.
介绍了文本分类的基本过程以及朴素贝叶斯和K近邻算法等基本分类方法,给出了基于覆盖的构造性神经网络分类算法,并将其与朴素贝叶斯和KNN作了实验比较。
A new method of modulation classification for digital communication signals,using normalized fourth-order cumulants as characteristics and a covering algorithm for constructing neural networks as a classifier,is proposed.
提出了一种把归一化四阶累量作为分类特征参数,应用神经网络覆盖算法进行分类的调制识别算法。
Its numerical value,average interval width,and empirical coverage are equal or similar to those of the Greenwood estimate for the variance of density function under various extreme clinica.
其数值、置信限平均宽度和经验覆盖在各种极端临床条件下均与Greenwood密度函数方差估计值相等或相近 ,而计算大大简化。