2022 Fiscal Year Final Research Report
Reservoir Self Organizing Map and Its Application
| Project/Area Number |
20K11992
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 61040:Soft computing-related
|
|---|
| Research Institution | Saga University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
中國 真教 福岡大学, 公私立大学の部局等, 准教授 (10347049)
|
|---|
| Project Period (FY) |
2020-04-01 – 2023-03-31
|
| Keywords | 時系列処理 / 自己組織化マップ / IoT機器 |
|---|
| Outline of Final Research Achievements |
To improve the performance of Reservoir Computing, a computationally inexpensive time series processing method, we developed an algorithm that classifies states using self-organizing maps and learns an output matrix for each state. The effectiveness of the algorithm was verified through experiments, and the algorithm was able to predict time series even in the unsupervised state (after learning), where the performance of conventional Reservoir Computing is significantly degraded. The effectiveness of this algorithm was also confirmed by applying it to the KDDCUP 2021 and 2022 data.
|
| Free Research Field |
ソフトコンピューティング
|
| Academic Significance and Societal Importance of the Research Achievements |
近年IoT機器の普及により、様々なデータが収集され、その解析や応用が行われているが、従来のAIの手法を用いて大規模な時系列データの処理を行うにはioT機器は計算能力が低く、また、ネットワークを用いてクラウドで処理するのにも、データ通信量が大きくなり、そのための電力消費が大きい、そこで、IoT機器などでも実行可能な、軽量で性能が良い時系列処理手法が求められ、その一つがReserviur Computing(RC)である、本研究課題では、自己組織化マップとRCを組み合わせたReservoir自己組織化マップを開発し、その有効性を実験により確認した。
|
