Machine learning analysis of Human Activity Recognition (HAR)

What's included

Within the download you'll find the following directories and files:

har/
├── data/
│   ├── clean_testing_data.csv
│   ├── clean_training_data.csv
│   └── HAR_data.zip
│        ├── HAR_training.csv
│        └── HAR_testing.csv
├── img/
│   ├── Confusion_Matrix_adelmo.png
│   ├── Confusion_Matrix_carlitos.png
│   ├── Confusion_Matrix_charles.png
│   ├── Confusion_Matrix_eurico.png
│   ├── Confusion_Matrix_jeremy.png
│   ├── Confusion_Matrix_pedro.png
│   ├── Variable_Importance_adelmo.png
│   ├── Variable_Importance_carlitos.png
│   ├── Variable_Importance_charles.png
│   ├── Variable_Importance_eurico.png
│   ├── Variable_Importance_jeremy.png
│   └── Variable_Importance_pedro.png
│
├── assignment.py
├── cleaning_data.py
└── plot_visualization.py

---

If you have a clean data, you don't have to first run script cleaning_data.py

Usage:

    $ ipython assignment.py

Results

carlitos
id class 9 A 11 B 18 B MSE = 0.03854

pedro
id class 1 B 17 A 19 B MSE = 0.03321

adelmo
id class 4 A MSE = 0.01969

charles
id class 10 A MSE = 0.00754

eurice
id class 5 A 13 B 16 E 20 B MSE = 0.03148

jeremy
id class 2 A 3 B 6 E 7 D 8 B 12 C 14 A 15 E MSE = 0.02644

Requirements

• Python 2.7+
• numpy, 1.8.2
• pandas, 0.17.1
• scikit-learn, 0.17
• matplotlib, 1.4.2

License