Iterative denoising (Q626230)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Iterative denoising
scientific article

    Statements

    Iterative denoising (English)
    0 references
    0 references
    0 references
    0 references
    0 references
    22 February 2011
    0 references
    Iterative denoising is a data mining technology for analysis of large heterogeneous datasets, e.g., sets of text documents. The result of it is a hierarchical divisive cluster tree with visual representation of each node. The specific feature of this technology is that the features for the clustering are extracted from data at each node of the hierarchy separately. This extraction is made by multidimensional scaling of the dissimilarities structure of a data set corresponding to the current node. The node division is made by k-means clustering. An application is presented to the analysis of 1047 texts from the Science News website.
    0 references
    cluster analysis
    0 references
    text mining
    0 references
    divisive hierarchical clusteriing
    0 references
    multidimensional scaling
    0 references
    k-means clustering
    0 references
    0 references
    0 references

    Identifiers