One subtask in constraint-driven placement is enforcing a set of orientation constraints on the devices being placed. Such constraints are created in order to, for example, implem...
Knowledge about forced moves enables to select a small number of moves from the set of possible moves. It is very important in complex domains where search trees have a large branc...
A power estimation approach is presented in which blocks of consecutive vectors are selected at random from a user-supplied realistic input vector set and the circuit is simulated...
We propose to selectively remove examples from the training set using probabilistic estimates related to editing algorithms (Devijver and Kittler, 1982). This heuristic procedure ...
We argue that K–means and deterministic annealing algorithms for geometric clustering can be derived from the more general Information Bottleneck approach. If we cluster the ide...