Classification
Build a function that maps $X$ to a categorical value $Y$.
- Linear classifier
- Nonlinear/Nonparametric classifier (Tree, Kernel, etc.)
Decision Tree
- http://www.cis.temple.edu/~giorgio/cis587/readings/id3-c45.html
Linear + various losses
- Logistic
- Hinge
- Naive Bayes
Choice of classifier
- If many categorical features: tree
- Real value feature: Logistic or SVM
- Tree: easy to overfit
- Logistic/Linear SVM: linear methods. weak classification strength
- Kernel SVM: slow to fit