The following project seeks to carry out a study of the electromyographic signals (EMG) coming from the MYO bracelet. The main characteristics of these will be acquired, then two artificial intelligence algorithms will be trained, in order to compare the most viable method according to the percentage of prediction for the design of the algorithm used and conclude which is the best method for classifying the movement. In the stage of validating the samples of a user on the neural network, it was possible to acquire up to 90% accuracy in the classification of a movement, this could be obtained through the confusion matrix.