[Note] from mit ai lab: how to do research

I. Reading

Read the paper: Find the most essential paper in the field. Read the list of suggestions in this field and obtain suggestions from senior users.

Read journals: the most valuable. AI: Artificial Intelligence (The Journal of Artificial Intelligence), Computational Intelligence, machine learning, ieee pami (Pattern Analysis and machine intelligence)

Three phases of reading the thesis:

1. Search for interest points

2. Find the true content points and highlights

3 Intensive Reading is necessary

Question: "how can I use this article?Article"," Is it really as the author claims? "," If ...... What will happen

Understanding the thesis, understanding the purpose of the thesis, the assumptions and choices made by the author, the problems related to the thesis, and the author's points of view.

If you are interested in this field, you should implement the demo in this paper.

2. Establish a relationship

Add secret paper pass network.

Have idea-> draft paper-> send it to a friend-> comments-> modify-> post

Frequently discussed

3. New fields of learning (physics, philosophy, linguistics, psychology ......)

Read textbooks

Read best journals

Read books from famous scholars

Four Notes

Record your thoughts, problems, and solutions

Regular reading

5. Writing

Organize ideas and write them into papers

Get Feedback

Method:

Outline in 1 column-> refined step by step

2. Write down all the things in the brain, even spam.

3. Write the core content first, not in sequence.

Give comments to others

Lecture 6

VII.ProgramDesign

Eight Psychological Factors

Failures and loss of interest often occur, and sometimes they are completely trapped and there is no progress.

Alan lakien's book "How to get control of your time and your life"

Persistence is generally more effective than tianzi.

Communication and feedback

Nobel Prize winners often doubt the correctness and value of their work.

(Original: http://www.cnblogs.com/jasonasm/archive/2004/09/01/38465.aspx)