This paper introduced a new feature selection method, which first used clustering to reduce redundancy among features and then used Information Gain to choose good features.
针对这一问题，提出了一种基于聚类的特征选择方法，先使用聚类的方法对特征间的冗余性进行裁减，然后使用信息增益的方法选取类别区分能力强的特征。
The reason of difference is analyzed and a new method named Combined Feature Selection is put forward.
分析了产生差异的原因，并提出一种适合于中文环境的特征选取方法：组合特征选取方法。
With regard to the problem that original feature in image classification is mass and redundancy, a new image feature selection method was presented.
针对目前图像识别中原始特征数量大、不相关特征多以及冗余等现象，提出了一种图像特征选择方法。
A new method of adaptive threshold selection is proposed which combines the edge feature of the color image based on the theory of mathematical morphology.
提出了一种自适应图像二值化方法，它基于数学形态学理论，求取彩色图像边缘特征，并将它与灰度直方图二值化方法相结合。
A new method of adaptive threshold selection is proposed which combines the edge feature of the color image based on the theory of mathematical morphology.
提出了一种自适应图像二值化方法，它基于数学形态学理论，求取彩色图像边缘特征，并将它与灰度直方图二值化方法相结合。
应用推荐