A Compression Technique for Non-Stationary Signals

In this book, we present a compression technique for non stationary signals such as Electroencephalography (EEG). We show that 90% compression is possible achieving very low reconstruction error. We show that the reconstructed compressed signals are suitable to use in applications such as seizure detection and Brain Computer Interface. We show a preliminary comparison of performance between the uncompressed and compressed signals of such applications. If you have any questions please contact the author directly. The code is available upon request.

Autorentext

Mr. Mahrous is a Data Scientist by training and practice holds an M.A.Sc. in pattern recognition and Artificial Intelligence from the University of British Columbia and a second M.A.Sc in the topics of Artificial Intelligence. He is also the founder and principle data scientist of Gauss-Insights.