2019 Fiscal Year Final Research Report
Development of an improved method for selective recording of execution history and its new applications
| Project/Area Number |
17K00096
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Software
|
|---|
| Research Institution | Kanazawa University |
|---|
Principal Investigator |
Sakurai Kohei 金沢大学, 電子情報通信学系, 助教 (80597021)
|
|---|
| Project Period (FY) |
2017-04-01 – 2020-03-31
|
| Keywords | 大規模データ処理 / アクターモデル / 機械学習 |
|---|
| Outline of Final Research Achievements |
In this study, we considered methods for large scale data processing, and we proposed and developed a method for data processing in tree models with the parallel and distributed environment using the actor model. Our method can cope with multiple models including online classification trees and hierarchical clustering with a design pattern which describes tree nodes as actors, and we showed that it can effiicently handle actula large scale data inputs from our experiments.
|
| Free Research Field |
プログラミング
|
| Academic Significance and Societal Importance of the Research Achievements |
本研究の成果は既存の機械学習アルゴリズムを大規模なデータにシームレスに対応させるための手法を提案している. 提案手法によりデータの分類やクラスタリングなどを扱うシステム開発が, 多くの開発者が慣れ親しんだ手法により理解しやすいモデルの定義によって迅速に行うことが可能となる. 結果としてデータの分析に関する多くの変更や性能の向上に関する要求に対応が容易となる.
|
