求翻译：针对DensityClust 算法中欧式距离测度不能充分反映复杂空间分布的数据集的结构特征。首先，我们提出了一种流形距离的计算方法，这种相似性测度方式能够更好地反映数据集的内在相似性。然后，根据密度的特点，一种新的密度峰值公式被重新定义，它是对参数具有不敏感性，并且能更好地代表局部密度。此外，为了能够自动选择聚类中心，提出了两阶段聚类中心测定方法。结果表明改进后的算法具有较高的精度和更好的鲁棒性与流形的分布、 多尺度和噪声重叠的数据。是什么意思？

DensityClust algorithm for Euclidean distance measure does not adequately reflect the complex structure of the spatial distribution of the data set.

For DensityClust algorithm central-distance measure does not fully reflect complex spatial distribution of data sets that characterize the structure. First of all, we have proposed a flow-distance calculation method, this Similarity Measure methods to be able to better reflect the data set of the si

Western-style cannot reflect the complex spatial distribution fully in view of the DensityClust algorithm in from the measure the data set structure characteristic.First, we proposed one manifold distance computational method, this similar measure way can reflect the data set well the intrinsic simi

DensityClust algorithm for Euclidean distance measure does not fully reflect the structural characteristics of the spatial distribution of complex data sets. First of all, we propose a manifold distance calculation method, this method of similarity measures to better reflect the intrinsic similariti

正在翻译，请等待...