目前,很多基于核密度估计的数据分类算法采用的判别规则忽视了不平衡类问题.对此,提出了改进的基于核密度估计的数据分类算法.该算法可处理不平衡类带来的影响,并在不平衡类问题严重时也能发挥好的效果,而且可以推广到多分类问题.实验结果表明了这种改进是非常有效的,它提高了基于核密度估计的分类算法对不平衡类的适应力. At present, many discriminant rules adopted by data classification algorithms based on kernel density estimation ignore the problem of unbalanced classes, and propose an improved data classification algorithm based on kernel density estimation, which can deal with the impact of unbalanced classes, And can also play a good effect when the imbalanced class of problems is serious, and can be extended to the multi-class problems.The experimental results show that this improvement is very effective, it improves the classification algorithm based on kernel density estimation of unbalanced class adaptability.