Study
Mythbusting data mining urban legends through large scale experimentation

Mythbusting data mining urban legends through large scale experimentation

Created 24-07-2016 by Martijn Post Visibility: public
Loading wiki
Data mining researchers and practitioners often use general rules of thumb or common data mining wisdom, those are so called data-mining myths. Even though, these myths are not always proven or sufficiently proven for general non-specific cases. Therefore, two data-mining myths are closely examined in this study. The two myths that were investigated are "Data preparation is more important than algorithm selection" and "Non-linear models perform better".