Download 699-approximation algorithm for Max-Bisection by Ye Y. PDF

By Ye Y.

Show description

Read or Download 699-approximation algorithm for Max-Bisection PDF

Similar algorithms and data structures books

Information, Randomness and Incompleteness: Papers on Algorithmic Information Theory: 008

The papers collected during this publication have been released over a interval of greater than two decades in greatly scattered journals. They resulted in the invention of randomness in mathematics which used to be awarded within the lately released monograph on “Algorithmic details conception” through the writer. There the most powerful attainable model of Gödel's incompleteness theorem, utilizing an information-theoretic method in response to the scale of computing device courses, used to be mentioned.

Introduction to Data Envelopment Analysis and Its Uses: With DEA-Solver Software and References

Advent to facts Envelopment research and Its makes use of: With DEA-Solver software program and References has been rigorously designed by means of the authors to supply a scientific creation to DEA and its makes use of as a multifaceted software for comparing difficulties in a number of contexts. The authors were occupied with DEA's improvement from the start.

Additional resources for 699-approximation algorithm for Max-Bisection

Example text

The gene of the first locus encodes the source node, and the gene of second locus is randomly or heuristically selected from the nodes connected with the source node (S) that is represented by the front gene’s allele. The chosen node is removed from the topological information database to prevent the node from being selected twice, thereby avoiding loops in the path. This process continues until the destination node is reached. Note that an encoding is possible only if each step of a path passes through a physical link in the network.

6. Convergence property of each algorithm. converging through smaller generations has better convergence performance because all the algorithms have the same population size in the experiment. In general, however, convergence performance must be compared with the average number of fitness function evaluations until the GAs reach equal quality of solutions [22]. A detailed explanation will be given in Sect. 2. , the optimal route), notwithstanding the somewhat inherent initial disadvantage. , networks types and scales).

The reason is not far to seek: the proposed algorithm involves the smallest number of fitness function evaluations. That means faster convergence. Networks with 15–50 nodes, and randomly assigned link costs were also studied. The results in respect of number of fitness function evaluations are shown in Fig. 8. , maximum difference is about 4%) is smaller than any other algorithm. In case of 30 nodes, for instance, the proposed GA is faster than Inagaki’s GA and Munetomo’s GA with prob. 2. Performance comparison on the rate of convergence.

Download PDF sample

Rated 4.61 of 5 – based on 32 votes