Multilayer Perceptron for Activity Recognition Using a Batteryless Wearable Sensor

International Conference on SMART CITY Innovation 2018

Putra, D.N.S.^1 ; Yulita, I.N.^1

Universitas Padjadjaran, Jl. Raya Bandung-Sumedang, Km. 21 street Jatinangor, Sumedang District

45363, Indonesia^1

关键词: Activity recognition;    Battery-less;    Classification approach;    MLP (multilayer perceptron);    Naive Bayesian;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/248/1/012039/pdf
DOI  :  10.1088/1755-1315/248/1/012039

来源: IOP

【 摘 要 】

Smart City is one of the trends around the world today. In order to achieve a smart city environment, we need everything to be connected and to be managed. As one of the aspects that have strong impact on people life is their own activities, a smart city should be accompanied by smart people which could be actualized by using sensors to recognize their activities. In this research, we present and evaluate a method to recognize the gesture of someone leaving bed using RFID device. We use a classification approach in our system to conduct the experiment. The method that we are using is MLP (Multi-Layer Perceptron). By using this method, we got 90.17% of accuracy, which is slightly better than Naive Bayesian that got 84.46% of accuracy.