Indoor Localization Using Machine Learning
Nowadays, smartphones can collect huge amounts of data in their surrounding environment with the help of highly accurate sensors. Since the combination of the Received Signal Strengths of surrounding Access Points and sensor data is assumed to be unique in every location, it should be possible to use this information to accurately predict a smartphone’s location. As it is very difficult to derive the correlation between these values, we must use machine learning methods. As part of this project, we have developed an Android application that is able to distinguish between rooms on a floor and special landmarks within a room. This has been accomplished using machine learning methods based on the Java library Weka. Ultimately, we hope to include this application into an indoor tracking system in order to improve its accuracy.
My thesis is available on https://github.com/JoelNiklaus/IndoLoc/blob/master/Bachelor_Thesis_Joel_Niklaus.pdf.
The documentation to my code is available as javadoc for most of the classes and methods directly in the code. Here there are some useful links for information about Android: https://developers.google.com/android/ and Weka: http://weka.wikispaces.com/, https://weka.wikispaces.com/Use+WEKA+in+your+Java+code,
Please see the file called INSTALL.
Please see the file called LICENSE.
I am happy to answer questions concerning my code under the following email address: joel.niklaus@students.unibe.ch.