Description
Wearable sensors have been shown to be effective for promoting self-awareness, wellness and re-education. For this project, you
will develop a detector for Body-rocking behavior from a blind subject (also commonly observed in individuals with autism) using
inertial measurements from a wearable system. Two sensors, one of the wrist and another on the arm, were used to record motion using accelerometers and gyroscopes.
The data will consists of sessions about 1-2 hours long each with annotations of when the behavior was observed. Your goal will be
to train a classifier for the detections of these events. This project will have two parts: In the first phase, we try
Random Forest, CNN, MLP which will serve as baseline for the second part of the project. For the second phase,we implement a recurring neural network (e.g., LSTM or CNN-LSTM) for learning features from the data and
improving on your results from the first part of this project.
Dataset
Dataset can be found [here](https://drive.google.com/drive/folders/1OKypsGFBXOTPibXxmCo0_ODiZe2d4l3e?usp=sharing)
Implementation
Implementation details can be found [here](Proj3a_report.pdf)
