Deep Learning for Cognitive Detection based on P300 Event-Related Potential

Due to the mechanical demands on real-time operations of brain signal reflection, manual parsing, and detection of the Event-Related Potential waveform, especially P300 evoked potential is impropriate and too slow. In addition, traditional methods require specialized knowledge and experience as well as experts to operate. Therefore, a faster and easier method is essential. The application of machine learning is a very reliable and popular method for classifying large and complex data in a short time. Therefore, this study presents a method to classify P300 in ERP waveforms using a deep learning algorithm which provides accurate and reliable results. It allows common and advanced users to detect the P300 evoked potential without specific skill and quickly deploy it to the BCI cases.