回头接着学习深度学习基础，推荐个很好的online book:http://neuralnetworksanddeeplearning.com/chap4.html,该章主要图形化的展示了为什么神经网络可以用来表示任何函数。有一句话很有意思，But it's worth reminding ourselves: the ability to compute an arbitrary function is truly remarkable. Almost any process you can imagine can be thought of as function computation.我所能够想象的任何过程都能被表示成为函数，什么样的问题不能表示为函数呢？也就是说什么样的问题不能用深度学习方法来解决呢？？回归到基础数学，什么是函数。来我们看看:
a metaphor of function
[for every element in input set(just collections, ex: a set of clothes), we have exactly one output element[single valued]. This is the difference between function and relation.]
interesting, Right?
A function can then be defined as a set of ordered pairs, "if it contains (a, b) and (a, c), then b must equal c"[single valued].
shocked and interesting?
