作者: Zhengwu Yuan , Xiaojian Zhang , Peng Zhou , Shanshan Wang
DOI: 10.1109/IICSPI48186.2019.9095921
关键词:
摘要: With the rapid development of wireless networks, positioning technology has also rapidly gained popularity. At same time, more and indoor scenes have received widespread attention. Therefore, high-precision is particularly important. This paper introduces model machine learning into location fingerprint positioning. Firstly, K-Nearest Neighbor (KNN) algorithm used to study accuracy location, compare it with performance Support Vector Machine (SVM), Random Forest (RF), Multi-layer Perceptron (MLP)algorithm in system. Finally, combined Particle Filter (PF) based on algorithm, experimental results show that particle filter improves accuracy, random forest highest accuracy.