Temporal Data Mining in Medicine and Bioinformatics ◾ 241
time series in progressively ner linear segments. Originally, the number
of linear segments is estimated using the number of cycles in the data set.
en a greedy approach that is based on a multiobjective function and uti-
lizes local and global information is used to estimate the linear segments.
In addition, breakpoints are estimated in the data set by measuring the
ability of each point in the data set to form bends.
In [Sat07], a medical time series clustering method is proposed, which
utilizes Formal Concept Analysis (FCA). FCA is a data representation
method, that identies conceptual structures in data sets and allows the
derivation of an ontology consisting of a concept ...