A group of representative measures of a set of time series and its decision support: a trial for dengue incidence data of Sri Lanka
Chekhaprabha Priyadarshanee Waduge, Naleen Chaminda Ganegoda, Darshana Chitraka Wickramarachchi, R. S. Lokupitiya, G. H. J. Lanel · International Journal of Information and Decision Sciences · 2022
Partitioning a time series into a set of subseries is required in some temporal investigations. Here, a cross-sectional approach is implemented for such subseries to design representative series. The series containing point-wise arithmetic means (mean series) is somewhat the simplest choice, when such a representative is required. However, it may not acquire important temporal variations. In this paper, several alternatives are proposed based on a class of measures called ultimate tamed series (UTS). Here, an operation called taming is implemented upon time series and it is non-commutative and non-associative. The taming is carried out with the aid of discrete Haar wavelet, which is subscribed by both point-wise concern and local trend given by an adjacent point. We conducted a trial for this using a dataset of disease incidence. Better-informed decisions are possible via the proposed architecture of processing time series data.