Taxicab non-symmetrical correspondence analysis
Abstract
En
Non Symmetrical correspondence analysis (NSCA) is a variant of the classical Correspondence Analysis (CA) for analyze two-way contingency table with a structure of dependence between two variables. In order to overcome the influence due to the presence of outlier, in this paper, it is presented Taxicab Non Symmetrical Correspondence Analysis (TNSCA), based on the taxicab singular value decomposition. It will show that TNSCA it is more robust than the ordinary NSCA, because it gives uniform weights to all the points. The visual map constructed by TNSCA offers a clearer perspective than the map obtained by correspondence analysis. Examples are provided.
Non Symmetrical correspondence analysis (NSCA) is a variant of the classical Correspondence Analysis (CA) for analyze two-way contingency table with a structure of dependence between two variables. In order to overcome the influence due to the presence of outlier, in this paper, it is presented Taxicab Non Symmetrical Correspondence Analysis (TNSCA), based on the taxicab singular value decomposition. It will show that TNSCA it is more robust than the ordinary NSCA, because it gives uniform weights to all the points. The visual map constructed by TNSCA offers a clearer perspective than the map obtained by correspondence analysis. Examples are provided.
DOI Code:
�
Keywords:
Non symmetrical correspondence analysis; taxicab singular values decompositio; $L1$ norm; robustness
Full Text: PDF
