Comparison of Linear and Tobit Modeling of Downscaled Daily Precipitation over the Missouri River Basin Using MIROC5

P-splines and GAMLSS: a powerful combination, with an application to zero-adjusted distributions

“…Response variables which can either be zero or a continuous positive quantity, are commonly encountered in diverse application areas. Some examples are: microbiome composition (Liu et al, 2019), daily precipitation and streamflow (Van Ogtrop et al, 2011; Popuri et al, 2015), annual insurance claim amounts (Heller et al, 2006), daily alcohol consumption (Liu et al, 2019), annual health services expenditure (Liu et al, 2019; Neelon et al, 2016b), annual expenditure on sports betting and online gambling (Houghton et al, 2019) and cattle mortality rates Belasco et al, 2012).…”

“…This article celebrates the merger of the two techniques by using an application in which the response variable has a zero-adjusted (or semi-continuous) distribution, that is, a mixed distribution, comprising a probability mass at zero and a continuous positive distribution (Rigby et al, 2019). These distributions have received attention in diverse application areas, such as medical (Liu et al, 2019), environmental (Popuri et al, 2015), actuarial (Heller et al, 2006), sports betting (Houghton et al, 2019) and agriculture (Belasco et al, 2012). Modelling approaches are generally either two-part models in which the zero probability and the continuous component are modelled independently, or a single model as in Tobit or Tweedie models.…”