基于局部估计密度的局部离群点检测算法

局部离群点检测是近年来数据挖掘领域的热点问题之一.针对交通数据去噪问题,提出一种基于局部估计密度的局部离群点检测算法,算法使用核密度估计方法计算每个数据对象的密度估计值,来表示该数据对象的局部估计密度,并在核函数的带宽函数计算中引入数据对象的k-邻域平均距离作为其邻域信息,然后利用求出的局部估计密度计算数据对象的局部离群因子,依据局部离群因子的大小来判断数据对象是否为离群点.实验表明,该算法在UCI标准数据集与模拟数据集上都可以取得较好的表现.

Local Outlier Detection Algorithm Based on Local Estimation Density

In recent years,local outlier detection is one of the hot issues in the field of data mining. In order to solve the problem of traffic data denoising,a local outlier detection algorithm based on local estimation density is proposed. The algorithm uses the kernel density estimation method to calculate the density estimation of the sample data which is used to represent the local estimation density.And the k-neighborhood average distance of the sample data viewed as its neighborhood information is introduced in the calculation of the bandwidth function of the kernel function. The value of the local outlier factor is calculated by using the local estimation density of the sample data,and is used to determine whether the data is an outlier. The experimental results show that the proposed algorithm can obtain better performance on both UCI datasets and real datasets.