Human Activity Analysis
Time series feature extraction for human activity monitoring
The objective of this project is to extract time series features for human activity monitoring.
First, I acquired the accelerometer data of all the 15 subjects for 4 activities: walking, running, climbing up and climbing down, from here. Next, I applied natural and horizontal visibility graph to the accelerometer time series data. Then, I carried out some network analysis such as computing average degree, network diameter, and average path length Permutation entropy and complexity are also computed.
Jupiter Notebook Link