Stage 1:
- collected 12 users single trail on straight/zigzag/stairs walk style, placement on foot/ankle/leg/thigh/lower back/wrist
- using python combine the data into one file
- c:\Python27\python CombineData.py input.csv output.csv UserId WalkingStyle
using WEKA to preprocess data
- load csv data file
- IRQ (extreme factor :6 ; outliner factor:3)
- Remove with Values: remove the outlier and extreme data(labeled with yes) we got 42437 lines of data(instances)
- Visualize them as follows:
- Using WEKA algorithm to identify the 12 users
| Algorithm | Correctly Classified Instances | Incorrectly Classified Instances | Root mean squared error | ||
|---|---|---|---|---|---|
| J48 -c 0.25 | 35763 | 84.2732 % | 6674 | 15.7268 % | 0.1589 |
- Using WEKA algorithm to identify the 3 actions
| Algorithm | Correctly Classified Instances | Incorrectly Classified Instances | Root mean squared error | ||
|---|---|---|---|---|---|
| J48 | 40784 | 96.1048 % | 1653 | 3.8952 % | 0.1552 |