ActionClassification

Using Acceleration and Rotationrate of a Device to estimate if sitting, running, biking or walking

Support Vector Classfier

Features used:

Trained/Tested with the PAMAP2 Dataset (subject 106) and got following accuracy:

with Tinkerforge IMU with TinkerLiveClassification.py (change to your UID)

Name		Name	Last commit message	Last commit date
Latest commit History 5 Commits
.ipynb_checkpoints		.ipynb_checkpoints
.gitignore		.gitignore
00_walking_Laufen_mit_Telefon_in_der_Hand.log		00_walking_Laufen_mit_Telefon_in_der_Hand.log
01_cycling_Radfahren_mit_Telefon_in_der_Hosentasche.log		01_cycling_Radfahren_mit_Telefon_in_der_Hosentasche.log
02_running_Joggen_mit_Telefon_in_der_Hose.log		02_running_Joggen_mit_Telefon_in_der_Hose.log
03_standing_Sitzen_mit_Telefon_in_der_Hand.log		03_standing_Sitzen_mit_Telefon_in_der_Hand.log
Classification.png		Classification.png
Classification.pptx		Classification.pptx
Classification2.png		Classification2.png
ClassifyActivity-pamap.ipynb		ClassifyActivity-pamap.ipynb
ClassifyActivity-pamap.py		ClassifyActivity-pamap.py
ClassifyActivity.ipynb		ClassifyActivity.ipynb
ClassifyActivity.py		ClassifyActivity.py
README.md		README.md
SVClassifier.pdf		SVClassifier.pdf
TinkerLiveClassification.ipynb		TinkerLiveClassification.ipynb
TinkerLiveClassification.py		TinkerLiveClassification.py
abs_acc_activities.pdf		abs_acc_activities.pdf
abs_acc_window.pdf		abs_acc_window.pdf
activitydata.csv		activitydata.csv
confusionmatrix.pdf		confusionmatrix.pdf
fft_amplitude_rotationrate.pdf		fft_amplitude_rotationrate.pdf
prepData.ipynb		prepData.ipynb
prepData.py		prepData.py