it will be one of the main aspects
of this paper.
2.3. Properties
The elements of a Fuzzy Cognitive Map are as follows:
Concepts: C1, C2, . . ., Cn. These represent the drivers and
constraints that are considered of importance to the issue under
consideration.
State vector: A = (a1, a2, . . ., an), where ai denotes the state of the
node Ci. The state vector represents the value of the concepts
/media/loftslag/Kok_JGEC658_2009.pdf
[Q(t);Q(t 1);Q(t 2)]
E[Q(ui);Q(ui 1);Q(ui 2)]
, ai =
Qb(t)
Qb(ui)
, bi =
Q(t) Qb(t)
Q(ui) Qb(ui)
,
and Qb(t) and Qb(ui) are baseflows calculated using the UKIH baseflow separation method
(Piggott et al., 2005). All rescaling coefficients were limited to a minimum value of 0.25 and a
maximum value of 5.
4.5 Deterministic prediction
A deterministic forecast was derived from the ensemble by taking
/media/vedurstofan/utgafa/skyrslur/2013/VI_2013_008.pdf