Evaluating the Accuracy of kNN Classifier for Gesture Detection Based on Forearm EMG Signal

The kNN algorithm is used to solve classification and regression problems. The idea and realization of this algorithm is one of the simplest in machine learning. Despite its simplicity, it solves the problem with a fairly high accuracy. When a person shows a gesture, an EMG signal is generated in the forearm, which can be captured using sensors. Based on these signals and gestures shown by the person, we train the classifier to recognize the gesture and send it to the microcontroller of the prosthesis. In this paper, with less execution time of the whole algorithm, it has been obtained 35% more accuracy than similar systems.