Now showing 1 - 3 of 3
  • Publication
    Clustering algorithm incorporating density and direction
    This paper analyses the advantages and disadvantages of the K-means algorithm and the DENCLUE algorithm. In order to realise the automation of clustering analysis and eliminate human factors, both partitioning and density-based methods were adopted, resulting in a new algorithm – Clustering Algorithm based on object Density and Direction (CADD). This paper discusses the theory and algorithm design of the CADD algorithm. As an illustration of its applicability, CADD was used to cluster real world data from the geochemistry domain.
    Scopus© Citations 7  908
  • Publication
    The application of cluster analysis in geophysical data interpretation
    A clustering algorithm which is based on density and adaptive density-reachable is developed and presented for arbitrary data point distributions in some real world applications, especially in geophysical data interpretation. Through comparisons of the new algorithm and other algorithms, it is shown that the new algorithm can reduce the dependency of domain knowledge and the sensitivity of abnormal data points, that it can improve the effectiveness of clustering results in which data are distributed in different shapes and different density, and that it can get a better clustering efficiency. The application of the new clustering algorithm demonstrates that data mining techniques can be used in geophysical data interpretation and can get meaningful and useful results, and that the new clustering algorithm can be used in other real world applications.
    Scopus© Citations 24  4617
  • Publication
    Research and application of clustering algorithm for arbitrary data set
    This paper discusses the theory and algorithmic design of the CADD (clustering algorithm based on object density and direction) algorithm. This algorithm seeks to harness the respective advantages of the k-means and DENCLUE algorithms. Clustering results are illustrated using both a simple data set and one from the geological domain. Results indicate that CADD is robust in that automatically determines the number K of clusters, and is capable of identifying clusters of multiple shapes and sizes.
    Scopus© Citations 5  1592