【摘要】 本文提出了一种新的有效的k近邻分类快速算法.该算法利用向量的方差和在小波域中的逼近系数得出两个重要的不等式.在搜索k近邻的过程中,首先判断每个训练向量是否满足这两个不等式,由此排除大量不可能成为k近邻的向量,从而可以快速的找到未知样本的k个近邻,使得在保持k近邻法分类性能不变的情况下,分类的效率得到很大地提高.最后,我们以纹理分类为例验证算法的有效性.更多还原

【Abstract】 A novel and efficient algorithm is proposed to reduce the computational complexi ty for KNN classification.It uses two important features,the approximation coef f icient of a fully decomposed feature vector with Haar wavelet and variance of th e corresponding untransformed vector,to produce two efficient test conditions.S i nce those vectors that are impossible to be the k closest vectors in the des ign set are kicked out quickly by these conditions,this algorithm saves largely the classification time and has the same classification performance as that of t he exhaust search classification algorithm.Experimental results based on texture image classification will verify our proposed algorithm.更多还原