Reviewing the current state of the data mining was found that this uses statistical algorithms, estimation or prediction of data and outputs in knowledge representation, which respond to a selection model to a data classification process to estimate new values for the selected attributes to model the dependencies according to the specific attributes, to displays the results of data mining and discovers new rules as a result of the extraction of data to achieve its applicability to certain objective. Almost most of the fields of know and knowledge is present data mining, that is like it find in the telecommunications, multimedia, education, trade and the financial sector, government management and medicine, extracting data for use in research, trends, habits, behaviors, predictions and estimates. That is why, it is essential that the selection of the use of different algorithms that meet condition of the problem of the which should to have a level of knowledge for make decision.