GT Capstone Spring 2016 - Team Paths
33 views · Published 24 May 2016 · 4:22 · Indexed 25 October 2026
Channel: Caroline Dunn · 2016 · Howto & Style
Team Paths developed a mobile app to provide recommendations based on your location, your preferences and your friend's preferences. They used a collaborative filter to feed into their recommendation system in the cloud (AWS). This mobile app collects individual information, running in the background and uses the least squares algorithm (similar to Netflix recommendations) to build the model. This team utilized geofencing on Android and iOS. They found that iOS was more accurate, and Android was significantly less accurate to the point where they used other background services to get meaningful location data.
