Open AccessA Machine Learning Approach to Interpolating Indoors Trajectories
Author(s) -
Daniel Rodrigues de Carvalho,
Daniel C. Sullivan,
Rafael Valladares de Almeida,
Carlos Caminha
Publication year - 2021
Language(s) - English
Resource type - Conference proceedings
DOI - 10.5753/kdmile.2021.17472
Subject(s) - computer science , latency (audio) , artificial neural network , artificial intelligence , random forest , real time computing , machine learning , object (grammar) , computer vision , telecommunications
In this article we propose a machine learning-based modeling to solve network overload problems caused by continuous monitoring of the trajectories of multiple tracked devices indoors. The proposed modeling was evaluated with hundreds of object coordinate locations tracked in three synthetic environments and one real environment. We show that it is possible to solve the problem of network overload increasing latency in sending data and predicting as server-side trajectories with ensemble models, such as the Random Forest, and using Artificial Neural Networks. We also show that it is possible to predict at least fifteen intermediate coordinates of the paths of the tracked objects with R2 greater than 0.95.