How important is data quality? Best classifiers vs best features

The task of choosing the appropriate classifier for a given scenario is not an easy-to-solve question. First, there increasingly high number algorithms available belonging to different families. And also lack methodologies that can help on recommending in advance family certain type datasets. Besides, most these classification exhibit degradation performance when faced with datasets containing irrelevant and/or redundant features. In this work we analyze impact feature selection over several synthetic and real experimental results obtained show significance selecting decreases after applying preprocessing step and, only alleviates choice, but it improves almost all tested.