Among the challenges in the field of medical sciences are those related to the classification of bacteria by taxonomy; However, with the emergence of vector support machines, it is possible to optimize this task through automatization by separating classes in space called hyperplanes.In this sense, the article assesses performance when classifying bacteria using the quadratic, cubic and radial Gaussian kernel functions.The results obtained allowed to conclude preliminarily that the VSM implemented in the research did not feature the best performance for the data sequence entered into the system, reaching a maximum global performance in the classifier no greater than 30.25%;However, it is necessary to continue applying modifications to the model developed in order to determine the possibility of increasing the success percentage.