In contrast with this general approach, Perretti and colleagues have
argued in a series of recent papers that model-free nonparametric
approaches may be a better way to tackle ecosystem complexity
(Perretti et al., 2013a,b,c). These nonparametric approaches rely on the
idea that one should be able to forecast future dynamics by analogy
with past dynamics. More precisely, a phase space representation of
past dynamics is used, in which variables of interest are plotted as a
function of each other, possibly with time delays (Kantz and Schreiber,
2004). Forecasts are then built using the collection of previously
observed dynamics, weighted by their proximity in the phase space
(see Section 2 and Perretti et al. (2013a,b) for additional details).