机器学习基石笔记 Lecture 1: The Learning Problem

Lecture 1: The Learning Problem

---

what is machine learning?

人类的学习：

机器学习：

why use machine learning?

机器具有自学习，相对于手工构造大量人工规则更可行，况且很多情况下，也没法人工建立规则

Give a computer a fish, you feed it for a day; teach it how to fish, you feed it for a lifetime. :-)

when use machine learning

首先得要有数据，其次要有内在规律，并且规律不便于人工抽取

Exercise

Application

Formalize the Learning Problem

$A$ takes $D$ and $H$ to get $g$

hypothesis

Relation

Machine Learning and Data Mining

difficult to distinguish ML and DM in reality

Machine Learning and Artificial Intelligence

ML is one possible route to realize AI

game tree 是啥?

Machine Learning and Statistics

statistics: many useful tools for ML

好吧，概率是干啥的？，我还真不是很清楚

Note:PLA每次修正后，按顺序找下一个错误的点，当找完一轮都没找着的时候，停下来。而Pocket，是随机找下一个错误的点，停止条件是达到迭代次数的上限