The 窶廚orrelat.窶 field is available only when the algorithm is chosen as 窶廚orrelation窶. In this you enter the correlation between the node and its predecessors, i.e., the nodes with edges pointing towards the current node. The correlation is a real number between -1 and +1.
A correlation close to -1 means that there is a very strong negative dependence between the node and its predecessors. That is, when the predecessors are small, the node tends to be large and vice versa.
A correlation close to +1 means that there is a very strong positive dependence between the node and its predecessors. That is, when the predecessors are small, the node tends to be small as well and vice versa.
If you specify a correlation close to 0, the node becomes more or less independent of its predecessors.
Note that it is generally not true that zero correlation implies independence. However, within the model framework applied by DynRisk, this statement holds.
