The BigMM2016 Challenge was completed without a first prize winner because no team meets the requirement of memory usage (64KB).
Recurrent Neural Network-Based Transportation Mode Detection
Toan Hong Vu
Advisor(s): Jia-Ching Wang
Media System Laboratory
Department of Computer Science and Information Engineering
National Central University, Taiwan
We present our approach of using Recurrent Neural Networks (RNNs) for detecting transportation modes in the BigMM2016 Challenge. That is an end-to-end system working efficiently with the provided data. Even though we only used data collected from accelerometer sensor, which leads to low power consumption, our model still achieves a remarkable accuracy.
Transportation Mode Classification Based on Decision Tree
Hao-Syuan Wang, Yu-Xaing Fei, Jheng-Yao Lin, Hao-Hsiang Liao, You-Chang Chung, Chun-Lin Chu, Yu-Che Cheng, Wen-Chen Lu, Huang-Chao Chan
Advisor(s): Shih-Hau Fang, Yu Tsao, Duan-Yu Chen
WIMOCLab & MISLab
Department of Electrical Engineering
Innovation Center for Big Data and Digital Convergence
Yuan Ze University, Taiwan
This work studies the transportation mode using big data from smartphone sensors. We use various machine learning algorithms to classify a smartphone user’s transportation mode and strike a balance between the accuracy, executive time, system delay, and model size. Finally, we choose a typical decision tree approach to do this project.