2018 Fiscal Year Final Research Report
A Quantum Bit Representation-Based Gene-Coding Method for Graph Optimization Problems and Evolutionary Computation Using the Method
| Project/Area Number |
16K00318
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Prefectural University of Kumamoto (2017-2018)
Ariake National College of Technology (2016) |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
飯村 伊智郎 熊本県立大学, 総合管理学部, 教授 (50347697)
|
|---|
| Research Collaborator |
NAKAYAMA Shigeru
|
|---|
| Project Period (FY) |
2016-04-01 – 2019-03-31
|
| Keywords | 量子風進化計算 / 進化計算 / 量子ビット表現 / グラフ最適化 / 最大カット問題 / 組合せ最適化 |
|---|
| Outline of Final Research Achievements |
To expand the applicable fields of the quantum-inspired evolutionary algorithm (QEA), we have proposed a gene-coding method that can represent graphs and have shown that the QEA implemented the gene-coding method can search approximate solutions through the experimental results using the maximum cut problem, which is one of the graph optimizations.
Furthermore, we have proposed a new measure that can estimate the state of the qubit for improving search performance. Introducing the proposed measure enables to maintain the diversity of the population and leads the search performance improvement.
We introduce a nonuniform rotation angle, which has various convergence speeds and is regarded as the individuality, into a quantum-inspired individual. Introducing the proposed individuality can eliminate the cumbersome process required to design a rotation angle while ensuring the quality of the obtained solution.
|
| Free Research Field |
ソフトコンピューティング
|
| Academic Significance and Societal Importance of the Research Achievements |
量子風進化計算手法の適用範囲を拡張し，グラフ最適化問題の一つである最大カット問題を用いた計算機実験によって近似解を探索できることを示した．厳密な最適解を求めることが困難な問題において，限られた時間の中で近似解を発見することは非常に有用である．
一方，確率振幅の収束状態を測定可能な指標は，効率的な解探索を実現し，非一様な回転角度を用いた個性は，解の探索性能を維持しつつパラメータ調整に係る煩雑な作業を軽減する．収束状態の測定指標および量子風個体の個性は，量子ビット表現を用いる進化計算手法であれば適用可能であり，最大カット問題を含むグラフ最適化問題だけでなく最適化問題全般への応用が期待できる．
|
