Duality can be exploited to solve problems, by considering simultaneously two perspectives-the primal and dual view of the problem.
These translations take the form of a duality: they turn argument-driven computation into demand-driven computation by exchanging input and output throughout, turning terms 'inside out'.
An interesting aspect of this duality is that it exchanges functional and imperative features.
Thus, in the proper topological setting for graphs that are not locally finite, duality extends in all its main aspects.
However, in general, duality is not yet well understood for problems involving constraints other than inequality constraints.
We propose taking advantage of the duality between prediction and compression.
This duality was often expressed in neighbouring sentences, providing an opening for selective quotation.
Joncieres argued earlier, for example, that 'duality' was a feminine trait.
