Читать книгу Data Analytics in Bioinformatics - Группа авторов - Страница 44

In this algorithm subsequent merging operations are performed on the dataset to arrange them into clusters. Dissimilarity coefficients are acquired using either Euclidean or Manhattan distance metrics [17]. These measured dissimilarities forms a matrix using unweighted pair group method with arithmetic mean [18].

Drawbacks of this method are that the objects once merged cannot be separated and different computation metrics leads to varying results.