From 2 billion in 2006, 200 billion devices are expected to be connected to the internet by 2020, each of which produces a different kind of data every day. An important part of this data consists of measurements that are captured from different sensors (e.g., gyroscope, compass and magnetometer). In addition, there exist different sensors (e.g., temperature, streetlight and parking) in our environment which plays an essential role in preserving the quality of our current lifestyle. Health-based sensory data (e.g., EEG, Diabetes and Cardiovascular data) can be considered as the most important data to be efficiently processed due to its importance in diagnosing diseases.
For this reason, it becomes more important to develop and use new techniques in order to understand these measurements and make use of them. In this project, two EEG (Electroencephalography) datasets will be provided in order to recognise the epileptic seizure and eye state. Different features and classifiers will be tested to achieve the highest possible accuracy, where a CNN model will be built. In the end, a different dataset (Human Activity Recognition using smartphone sensors) will be provided to test the generality of the model and the selected features.
Therefore, two teams (4-5 students each) will be formed to work independently. To register, send a request with your name, registration number and your course of study to firstname.lastname@example.org