Study on Machine Learning of Health Big Data Based on Cloud Computing
Project/Area Number |
26350868
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Applied health science
|
Research Institution | Takasaki University of Health and Welfare |
Principal Investigator |
|
Co-Investigator(Kenkyū-buntansha) |
倉林 正彦 群馬大学, 大学院医学系研究科, 教授 (00215047)
児玉 直樹 高崎健康福祉大学, 健康福祉学部, 准教授 (50383146)
|
Project Period (FY) |
2014-04-01 – 2017-03-31
|
Project Status |
Completed (Fiscal Year 2016)
|
Budget Amount *help |
¥4,810,000 (Direct Cost: ¥3,700,000、Indirect Cost: ¥1,110,000)
Fiscal Year 2016: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2015: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2014: ¥2,340,000 (Direct Cost: ¥1,800,000、Indirect Cost: ¥540,000)
|
Keywords | 健康データマイニング / 個人健康管理システム / 時系列データ解析 / クラウドコンピューティング / 機械学習 / 遅延相関分析 / 健康ビッグデータ / 健康データマニング / 季節変動補正 / 末梢血管血流速度計測 / 血液循環 |
Outline of Final Research Achievements |
Long-term time-series data of an individual’s home blood pressure measurements were analyzed to clarify the relation with lifestyle components such as energy expenditure and supply. We found that the distribution of daily blood pressure data approached normal when the seasonal variation observed in the time-series data was adjusted. This demonstrates that adjustment is desirable for producing appropriate rules between blood pressure and lifestyle components in the healthcare data mining process. The results of healthcare data mining using the adjusted data are summarized as follows:(1)The adjustment of blood pressure data affected the rule production process and different rules were extracted with the adjustment because the blood pressure data were categorized into three classes (higher, moderate, or lower) when being used for an output (target) variable.(2)Rules extracted with the adjustment were more useful for personal healthcare than rule extracted without adjustment.
|
Report
(4 results)
Research Products
(12 results)