I used Python to implement digital signal processing methods, outlier detection and temporal averaging to clean the recorded signals. I then used dimensionality reduction and unsupervised learning to identify signals corresponding to a subjects' hand opening and closing and some sign-language letters.