Abstract The system architecture presented in Chapter controls (i.e., sustains and constrains) the invocation of inference methods introduced Chapter. In this chapter, we describe higher level more detail.

Compositional data are nonnegative carrying relative, rather than absolute, information—these often with a constant-sum constraint on the sample values, for example, proportions or percentages summing to 1% 100%, respectively. Ratios between components of composition important since they unaffected by particular set chosen. Logarithms ratios (logratios) fundamental transformation in ratio appro...

Trend filtering is known as the technique for detecting piecewise linear trends in univariate time series. This extended to setting of compositional data, which are multivariate data where only relative information importance. According this, we formulate problem and present a procedure how efficiently solve it. To show usefulness this method, consider number COVID-19 infections several Europea...