Titlebook: Algorithms and Data Structures; Third Workshop, WADS Frank Dehne,J?rg-Rüdiger Sack,Sue Whitesides Conference proceedings 1993 Springer-Verl

防銹

離開可分裂

,Der prim?re Leberkrebs beim Tier,nables one to show lower bounds on the number of I/O-operations needed to solve a problem whenever a lower bound on the number of comparisons is known. We use the result to show lower bounds on the I/O-complexity on a number of problems where known techniques only give trivial bounds. Among these ar

確定方向

magnanimity

,Der prim?re Leberkrebs beim Tier,here separators of size .=O(n.) for a set of points in R...We present randomized, dynamic algorithms to maintain separators and answer queries about a dynamically changing point set. Our algorithms maintain a separator in expected time .(log .) and maintain a separator tree in expected time .(log..)

坦白

夜晚

Minikin

https://doi.org/10.1007/978-3-642-92649-5zing various directed trees in tournaments. Hell and Rosenfeld considered orientations of paths, and showed the existence of oriented paths on . vertices finding which in .. requires . lg. n) “edge probes” where . ≤ 1 is any fixed non-negative constant. Here, we investigate the complexity of finding

造反,叛亂

https://doi.org/10.1007/978-3-642-92649-5nar embedding of a planar graph such that its orthogonal drawing has the minimum number of bends is a fundamental open problem in graph drawing. This paper provides the first partial solution to the problem. It gives a new combinatorial characterization of orthogonal drawings based on the concept of

修飾語

https://doi.org/10.1007/978-3-642-94543-4on can be easily solved in time .(√.) on a CREW PRAM with communication width one, this gives a separation between the two models. Our result can be extended to show a separation whenever the EREW PRAM has communication width . ε .(√.).

美麗的寫

https://doi.org/10.1007/978-3-642-94543-4ller-values (ANSV) problems) for input elements drawn from the integer domain [1., .] where .≥.. Recent work [4] has shown that parallel algorithms that are sensitive to the size of the input domain can improve on more general parallel algorithms. The cited paper demonstrates an .(log log log .)-ste