Predict Your Location with WiFi and Machine Learning
Summary of my bookmarked Github repositories from Apr 10th, 2017
"whereami" is a cross-platform Python package that utilizes WiFi signals and machine learning (sklearn's RandomForest) to predict your location, even for short distances of 2-10 meters. It works on OSX, Windows, and Linux (tested on Ubuntu/Arch Linux). The package "access_points" allows scanning WiFi in a cross-platform manner and provides JSON output. You can use "whereami" to learn and label different locations, perform cross-validated accuracy checks, and predict your current location. The package is highly accurate, especially for distances of 10 meters or more, and can distinguish between different locations within close proximity.