Semiparametric Modelling of Environmental Data via Generalised Sigmoidal Curves

Many natural phenomena, like bacterial or plant growth, epidemic contagions and some diseases' progression, present patterns that are sigmoidal or referable to sigmoidal: i.e. they present one or two inflection points where both the curve trend and concavity change. The traditional approaches to model this kind of curves are parametric, but their formalisation is often too rigid. In contrast, nonparametric models show excellent fits, but are not interpretable. In this work, we propose a `compromise' between these two paradigms, i.e. a semiparametric approach, which ensures model flexibility as well as interpretable parameters, including inflection points. We show our proposal's ability to model nonstandard data through some applications to environmental data.