| Budget Amount *help |
¥2,990,000 (Direct Cost: ¥2,300,000、Indirect Cost: ¥690,000)
Fiscal Year 2017: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2016: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2015: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
|
|---|
| Outline of Final Research Achievements |
We identified 40 subjects using some machine learning algorithms based on touch operations history when they performed basic operation, text browsing, and web browsing on our Android application. The results showed that user identification rate reached about 95% for basic operation and text browsing. In addition, we recorded 10 touch operations histories of 11 subjects for a half year to examine the long-term changes. We confirmed that the correctly classified rates for pinch gestures and text browsing are almost constant as the number of experiments increases.
From the 3-axes accelerometer data of 15 subjects when they walked not only on flat ground but also ascend and descend stairs, we extracted 52 features adding new features to the previous 43 features, and then selected the subset of the 52 features with small number and high accuracy. We confirmed that the accuracies of going upstairs and downstairs are improved by the feature selection.
|
