Application of Machine Learning to GNSS Data Collected by Smartphone
Trends showing increase in the number of mobile device users, as well as the number of tourists, imply that more people rely on their smartphones when navigating in a new environment. Based on these facts, the idea for this experimental research appeared. That idea is applying the process of machine learning, more precisely, the implementation of a neural network, to investigate the possibility of improving the accuracy of smartphone navigation. The achieved results indicate that machine learning algorithms (neural networks) are a powerful tool that can also be applied to GNSS data collected by a smartphone device, in order to improve accuracy. Based on the collected data in the field, preprocessing and machine learning process, it is concluded that it is possible to improve the accuracy of mobile device navigation by up to 50%.