Читать книгу Data Analytics in Bioinformatics - Группа авторов - Страница 48

In this algorithmic approach entire data in the dataset is maintained in the form of a tree structure called as CF (clustering Feature) which is based on clustering feature. The CF tree also holds the information of clustering. As the outliers are eliminated the sub clusters data is made as leaf nodes. With iterative updating of CF BRICH provides the potential to deal with huge datasets. This works effectively for gene expression data [22].

Drawback—This algorithm gets skewed with the presence of nonspherical clusters [22].