Nerds

Tonight one of my friends said that she's the biggest nerd that she knows. As a rebuttal, I'd like to present the first few lines of a paper that I turned in for a class yesterday:

Decision trees are simple models for multi-class classification problems. A decision tree classifies a vector of mixed discrete and/or continuous features by examining the most discriminative feature at each step, narrowing down the classification possibilities until only one remains.

Decision trees can be inferred from labeled data using supervised machine learning. Trees can optionally be pruned to correct for overfitting and thus improve their ability to generalize to unseen data.

So there. :)