Preliminary results of evaluating a prediction-based algorithm for movement pattern recognition and classification

In this paper a new, easy-to-use algorithm is presented to support telerehabilitation of people with movement disabilities. This algorithm can adapt to the needs of the patients as it combines the cognitive property of intelligent decision-making systems with the human-computer interaction-based movement therapy. This algorithm is built upon the earlier work of the authors and it uses six various mean techniques to classify gesture descriptors. The preliminary results and a plan of evaluating the algorithm are presented in this paper.