Hand Movement Classification using LDA, K-NN, and SVM

Brain-computer interface (BCI) is an artificial intelligence system that permits the control and communication between the human brain and a machine or an external device by acquiring a subject's brain signals using the electroencephalogram (EEG) and converting the electrical activity into meaningful information for the machine. This interface is designed to improve the lives of people who have lost motor function to regain their independence and quality of life. EEG stands out as the most efficient and widely used measurement method, due to its non-invasive nature, portability, and affordability. The objective of this work is to classify left and right-hand movements performed by three subjects based on the brain activity of motor imagery with the dataset BCI competition IIIb. For that, statistical features are used and different linear discriminant analysis (LDA), k-nearest neighbors (KNN), and support vector machine (SVM) algorithms of classification are used. In the findings, the SVM method showed the highest classification accuracy of 76.9%.