Selective neural network in training set on indoor positioning algorithm

Abstract

We propose a new method to combine the Knn algorithm with the neural network. Due to the size of the training set affects the accuracy of the training results, we propose a neural network with selective training set. Its main method is changed the size of the training set by a control variable to get the most suitable training data. The advantage of this method is that the training set can be adjusted to achieve a trade-off between the positioning accuracy and the training set size. The result of Knn algorithm positioning is more accurate by adding BP (Back Propagation) selective neural network algorithm.