Three Types of Learning

监督学习的两个分类：回归和分类 

Unsupervised learning AIM:

    -create an internal representation of the input that is useful for subsequent supervised or reinforcement learning

    -It provides a compact ,low-dimensional reqresentation of the input.高维数据往往存在或者近似于一个低维流形（low-dimensional manifold）(PCA)

    -It provides an economical high-dimensional representation of the input in terms of learned features.

            -Binary features are economical(每个变量只需要一个比特来说明其状态)

            -So are real-valued features that are nearly all zero.

-It finds sensible clusters in the input.聚类是非常稀疏的编码，对于每个聚类我们有一个变量。