Skip to content

ETH Zurich Master's Thesis "Deep vs. Classical Machine Learning Approaches for Transportation Mode Recognition in a Digital Health Application" for Classifying Bus/Walk/Cycle/Car/Tram/Train with up to 94% accuracy.

Notifications You must be signed in to change notification settings

naefl/transportation_mode_classifiers

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

transportation_mode_classifiers

ETH Zurich Master's Thesis "Deep vs. Classical Machine Learning Approaches for Transportation Mode Recognition in a Digital Health Application" for Classifying Bus/Walk/Cycle/Car/Tram/Train with up to 94% accuracy based on Android Sensors and GTFS information from Zurich.

Compares SVMs, Boosted Trees and others with RNNs (GRU/LSTM). Includes Dataparser and feature calculation script. Includes param optimization and feature selection. Includes KDA/LDA based dimensionality reduction.

Sensors used: GPS, Accel, Gyro, Sound Pressure.

Main frameworks: sk-learn, TensorFlow 1.7

About

ETH Zurich Master's Thesis "Deep vs. Classical Machine Learning Approaches for Transportation Mode Recognition in a Digital Health Application" for Classifying Bus/Walk/Cycle/Car/Tram/Train with up to 94% accuracy.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages