This paper considers the wireless indoor localization problem and aims at improving the localization precision for advanced medical and industrial applications. We propose an enhanced machine learning approach based on the support vector regression (SVR) model, with the received signal strength indicator (RSSI) values as features. The proposed method incorporates the ensemble learning concept and a new weight assignment scheme for ensemble learning. Numerical results demonstrate improved prediction performance of the proposed scheme in comparison with common machine learning techniques for localization such as artificial neural network (ANN), k-nearest neighbor (k-NN), and conventional SVR in simulated indoor environments.
High-Precision Wireless Indoor Localization via Weight-Learning Ensemble Support Vector Regression
2015-09-01
284815 byte
Conference paper
Electronic Resource
English
Novel Selective Support Vector Machine Ensemble Learning Algorithm
British Library Online Contents | 2008
|Short-term metro passenger flow forecasting using ensemble-chaos support vector regression
Taylor & Francis Verlag | 2020
|A Comparative Analysis of BLE-Based Indoor Localization with Machine Learning Regression Techniques
Springer Verlag | 2024
|