Learning with Different Output Space $Y$
- binary classification: $Y=\{-1,+1\}$
- multiclass classification: $Y=\{1,2,…,K\}$
- regression: $Y=\mathbb{R}$
- structured learning: $Y=structures$
Learning with Different Data Label $y_n$
- supervised: all $y_n$
- unsupervised: no $y_n$
- semi_supervised: some $y_n$
- reinforcement: implicit $y_n$ by goodness($\tilde{y}_n$)
Learning with Different Protocol $f\Rightarrow (x_n,y_n)$
- batch: all known data
- online: sequential (passive) data
- active: strategically-observed data
Learning with Differeent Input Space $X$
- concrete: sophisticated (and related) physical meaning
- raw: simple physical meaning
- abstract: no (or ittle) physical meaning
Focus: [classification/regression] [supervised] [batch] [concrete]