[1] Machine Learning Notes-Introduction
Intro
Recently, I was preparing for my undergraduate thesis/design, which is concerned with quadruped robot. However, I find my design and researches never deepen down the key point of algorithm. Thus, I was puzzled about what I learned during undergraduate: Is Mechanical Engineering we learned before fit the trend of the world? Need I prepare to transfer my research area? Then, I was rethinking my career and interest fields, as a result, I was stuck in the infinite thinking and anxiety.
Even worse, a very bad news also push a dilemma against me.
U.S. to Expel Chinese Graduate Students With Ties to China’s Military Schools
Would I be permitted to keep graduate studying in US? Should I find a job instantly? No one could give me an answer now, but Warald released some good analytical passages in his website, he said:
和平年代，成长是线性的；风云迭起，才是指数级的.
Good news is that my college is beyond the list which provide reference to US government which universities are “risky” for US people. And my project is not concerned with military. However, the uncertain future would still make me worried about my VISA.
Worry is a spasm of the emotion, the best way to get out of it is doing something I like. Fortunately, I like my major (by now). So, instead of worrying about my VISA, I better utilize these time to learn some knowledge. So Machine Learning comes to me and I decided to learning ML course by Andrew Ng which is a pretty good online course in my mind. You can take this course on coursera.
The aim to write this passage is to backup my notes and build a structure/mind map of knowledge. This passage also includes some key points of algorithm\mistake during my studying.
What is Machine Learning ?
Prof. Tom Mitchell from CMU provides a modern definition:
A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.
To make this tongue twister clear, I make some explanation about TEP:
T means Task—-what u want to do
E means previous experience
P is like some feedbacks, how many are correct
Take Email Spam as an example:
| definition | example | |
|---|---|---|
| T | Task | classify email as spam/ not spam |
| E | Experience | know your previous labeled emails as spam/not spam |
| P | Performance/feedbacks | the number of emails are correctly classified as spam/not spam |
This would be more intuitive to understand the basic definition of Machine Learning.
Generally, we can classify ML into two parts: Supervised learning and Unsupervised learning. Some people also think deep learning should be included, the picture below could easily explain inner connection between AI, Machine Learning and Deep Learning.
from left to right : AI, Machine Learning, Deep Learning.
I hope I could apply ML to practical ME projects and maybe it can promote my competitiveness in market. So I will keep writing memos to record my steps on studying ML.
You can download xmind map files using links here. code：304
By the way, I really like a quote mentioned by Prof.Ng:
Science is NOT a battle, it is a collaboration. We all build on each other’s ideas. Science is an act of love, not war. Love for the beauty in the world that surrounds us and love to share and build something together. That makes science a highly satisfying activity, emotionally speaking!
