Class TrimmedMeanApproach<N>

  • Type Parameters:
    N - Neighborhood object type
    All Implemented Interfaces:
    Algorithm, OutlierAlgorithm

    @Title("A Trimmed Mean Approach to Finding Spatial Outliers")
    @Description("A local trimmed mean approach to evaluating the spatial outlier factor which is the degree that a site is outlying compared to its neighbors")
    @Reference(authors="T. Hu, S. Y. Sung",
               title="A trimmed mean approach to finding spatial outliers",
               booktitle="Intelligent Data Analysis 8",
               url="http://content.iospress.com/articles/intelligent-data-analysis/ida00153",
               bibkey="DBLP:journals/ida/HuS04")
    public class TrimmedMeanApproach<N>
    extends AbstractNeighborhoodOutlier<N>
    A Trimmed Mean Approach to Finding Spatial Outliers.

    Outliers are defined by their value deviation from a trimmed mean of the neighbors.

    Reference:

    T. Hu, S. Y. Sung
    A Trimmed Mean Approach to finding Spatial Outliers
    Intelligent Data Analysis 8

    the contiguity Matrix is definit as
    wij = 1/k if j is neighbor of i, k is the neighbors size of i.

    Since:
    0.4.0
    Author:
    Ahmed Hettab
    • Field Detail

      • LOG

        private static final Logging LOG
        The logger for this class.
      • p

        private double p
        the parameter p.
    • Constructor Detail

      • TrimmedMeanApproach

        public TrimmedMeanApproach​(NeighborSetPredicate.Factory<N> npredf,
                                   double p)
        Constructor.
        Parameters:
        p - Parameter p
        npredf - Neighborhood factory.
    • Method Detail

      • getInputTypeRestriction

        public TypeInformation[] getInputTypeRestriction()
        Description copied from interface: Algorithm
        Get the input type restriction used for negotiating the data query.
        Returns:
        Type restriction
      • run

        public OutlierResult run​(Database database,
                                 Relation<N> nrel,
                                 Relation<? extends NumberVector> relation)
        Run the algorithm.
        Parameters:
        database - Database
        nrel - Neighborhood relation
        relation - Data Relation (1 dimensional!)
        Returns:
        Outlier detection result