Classifier_ranking

Part of the class of Computational Techniques for Machine Learning. The activity consists of evaluating and analyzing the results of the algorithms BRM, GMM, ISOF, and ocSVM according to AUC, without scaling or normalizing the database, using MinMax scaling, and using Standard normalizing, in 60 different databases provided by the professor. Visualization is done with a box plot, the Friedman test with a posthoc test, a CD diagram with the results of a statistical test, and the obtained results of the evaluated algorithms are discussed. The BRM implementation was modified, with at least three dissimilarity measures.