摘要：AbstractThis paper shows the results of a proposal to give orders to a drone using electroencephalographic signals, which are filtered between frequencies of 8 and 13 Hz, also some preprocessing in three different types of classifiers are evaluated: random forest, nearest neighbors and convolutional. The best result was the convolutional type network without any preprocessing and with an accuracy higher than 80% and an average time response of 42.21 ms.
