Indoor-positioning of wireless devices by using data mining algorithms with 2 access points

Abstract

This research aims to study wireless device indoor positioning methods by using machine learning algorithms, i.e; Decision Tree, Naive Bayes, Artificial Neural Networks, and K-Means by exploiting signal strength from 2 access points. The performance comparison is done in terms of accuracy of classification of positions, precision of distance classified, complexity of distinguish modeling, and effects of classification of positions on results from quantity of learning data in order to find the suitable algorithm. The result of this study can suggest that the Decision Tree algorithm and the Naive Bayes algorithm are suitable for future indoor-positioning software development.