a point has a clear pattern to join the tree at different stages with

data points. Therefore the relationship between data points can

investigated. The hierarchical clustering algorithm is such an

m.

e hierarchical cluster analysis algorithm

archical cluster analysis algorithm merges data points gradually

nging tree [Jardine and Sibson, 1971]. Afterwards, such a tree is

nerate subtrees if it is required. Each of the cut subtrees is treated

er. The most attractive feature of the hierarchical cluster analysis

m is that the pairwise relationships between data points can be

sualised in a hierarchical structure.

ierarchical cluster analysis algorithm has been used to classify

s of two varieties using phenolic compounds analysis and

ed two major clusters of red wines [Stoj, et al., 2020]. In a colon

tudy, it has been used to evaluate different approaches and

that Raman spectroscopy shows the best performance for colon

agnosis [Depciuch, et al., 2020]. The hierarchical cluster analysis

m has also been used to discover extremely small volume of

g molecules and to separate molecules based on bacteriophage-

orimetric sensor array technology [Kim, et al., 2020]. In blood

e subtype research which aims to discover biomarkers for

scular inflammation, the knowledge of human monocyte

neity is still imperfect. Therefore, a study has employed the

cal cluster analysis algorithm to discover circulating human

e subtypes based on expression data provided by high-throughput

ometry screening arrays [Hoffmann, et al., 2020]. The study has

ed 242 specific markers and revealed some new markers which

e classical, intermediate and non-classical subtypes.

n a data set D, the algorithm finds a pair of data points (ܠ௡∈ࣞ

∈ࣞ ) with the least distance or the maximum correlation

nt at one cycle of a clustering process and thus merge them. The