From MIT Artificial Intelligence Lab: How to do research?

xiaoxiao2021-03-06  21

· Massachusetts Institute of Technology

Artificial intelligence laboratory

Ai Working Paper 316 October 1988

From MIT Artificial Intelligence Lab: How to do research?

Author: AI Lab graduate students all

EDTOR: David Chapman

Version: 1.3

Time: September 1988

Translator: Liuquan Bo Beijing Normal University Information College 2000 PhD

Abstract The main purpose of this paper is to explain how to do research. We provide these suggestions to do research itself (reading, writing, etc.

Program design, understanding the research process and starting love research (methodology, topic, selection, emotional factors),

It is very worthless.

Copyright 1987, 1988 Author Copyright

Remarks: Working Papers of Artificial Intelligence Lab for internal communication, including information or over

Can't publish in detail. Unlike formal papers, all references will be listed.

1 Introduction

what is this?

There is no goddess medicine to ensure success in the study, this article only lists some may help.

Informal comments.

Who is the target reader?

This document is mainly written for graduate students in the MIT Artificial Intelligence Lab, but the artificial intelligence of other institutions

It is also very valuable. Even the researchers in the artificial intelligence, you can also find some part of yourself.

how to use?

It is too long to read this article, it is too long, it is best to use the way. Many people think that the following methods are very effective: first

Read it quickly, then select the part of the part that is related to your current research project.

This document is roughly divided into two parts. The first part involves various skills required to researchers: reading, writing and way

Order design, etc. The second part discusses the research process itself: What is going on? How to do research, how to choose

Questions and Church, how to consider emotional factors in the study. Many readers reflect, from the long run, the second part is higher than the first

More valuable, more interested.

• How to make the foundation of AI research by reading. Listed important AI journals, and give some reading

.

? How to become a member of the AI ​​research field: keep in touch with relevant personnel, they can keep you before the study

Tracking along the edge know what material should be read.

? Section 4 Learn knowledge in the relevant fields of AI. There are basic understanding of several fields, for one or two areas

proficient.

? How to make research notes.

How to write journal papers and graduation thesis. How to write reviews for drafts, how to use others's review

. How to publish a papers.

• How to do research reports in the section 7.

• The section 8 is related to the program. The AI ​​program design is different from the process of usual habits.

? The section 9 is the most important issue of the research workfare, how to choose a tutor. Different mentors have different styles, this section

Comments help you find the right mentor. The tutor is you must understand how to use resources.

? Size 10 About the graduation papers. Graduation thesis will occupy most of the graduate career, this part involves how to select questions

And how to avoid waste time.

? The section 11 is related to research methodology, has not been completed.

• Section 12 is the most important section: Emotional factors in the research process, including how to face failed, how to set up

Target, how to avoid insecurity, keep confidence, enjoy happiness.

2. Read

Many researchers spend half a time reading literature. You can learn a lot quickly from others' work. This section is discussed

On the reading in the AI, the fourth section will discuss other subjects related reading.

Read the literature and start on today. Once you start writing a papers, there is not much time, and the reading is mainly concentrated in the literature related to the theme of the paper. In the first two years of graduate students, most of the time should be used to do course work and play foundations. this

Reading textbooks and publishing journal articles can be. (In the future, you will mainly read the draft of the article, see the section

three).

The reading amount required to make a solid foundation in the field is a persistent. But since AI is just a small

In the field of research, you can still spend a few years to read the most nature of many papers published in the field.

section. A useful little skill is first to find out the most essential papers. You can refer to some useful bibliography:

For example, graduate courses, other schools (mainly the Stanford University) Postgraduate admission list, this

Some can make you have some initial impression. If you are interested in a sub-domain of Ai, to the high grade of this field

What is the most important ten papers in the field of this field, if you can, borrow a copy. Recently, there have been a lot of care

Edited the papers on a sub-field, especially Morgan-Kauffman published.

AI laboratory has three internal publications series: Working Papers, Memos and Technical Reports, formal

The degree increases, and can be found on the eight-layer shelf. Review the publications in recent years, I am very interested

Copy it. This is not only because many of them are very significant papers, but also to understand the progress of laboratory members.

It is also very important.

There are a lot of journals about AI, fortunately, only part of it is worth seeing. The core journal is Artificial I

NTELLIGENCE, there is also written "The Journal of Artificial Intelligence" or "Aij". Ai collar

The domain that truly valuable value will eventually be cast to Aij, so it is worth browsing the AIJ of each period; but the journal

There are also many papers to make people feel annoying. Computational Intelligence is another journal worth watching. Cogn

Itive Science also publishes a lot of significant AI papers. Machine Learning is the most important in the field of machine learning.

Resource. IEEE PAMI (Pattern Analysis and Machine Intelligence is the best related vision

Journals, each period has two or three valuable papers. International Journal of Computer Vision (IJ

CV) is the latest founder, it is still worth it. Robotics Research's article is mainly about motivation

Sometimes, sometimes there is an epoch-making intelligent robot papers. IEEE ROBOTICS and Automation occasionally good text

chapter.

I should go to the computer science library of the school (one floor of the Tech Square of MIT), flour over other hospitals.

The school published AI technology report, and select carefully reading yourself.

Reading the paper is a skill that needs to practice. It is impossible to read all the papers completely. Reading papers can be divided into three stages:

The first stage is to see if there is something interested in the paper. The AI ​​paper contains a summary, which may have content, but

It is also possible that there is no or concluded that it is not good, so you need to jump, this look at it, understand the author

What is it. Content Directory, Conclusion Part (Conclusion) and Introduction (INT

Roduction is three key points. If these methods can't, you have to browse quickly. Once you understand the probably and innovation points of the papers, you can decide whether it is necessary to carry out the second phase. In the second phase, you must find out the truse of the paper

There is a part of the content. A lot of 15 pages can be rewritten as a page of about a page; therefore, you need to find those true

Invisible place, this is often hidden in a place. The author is interested in its work,

It is not always interested in, and vice versa. Finally, if this paper does value, return to the reading.

Keep in mind a question when reading the papers, "How should I use this pacription?" "Really like the author claims?"

"If ... what happens?". What conclusions in understanding the paper are not equivalent to understanding the paper. Understanding

Text, you must understand the purpose of the paper, the author's choice (many are implicit), whether assumptions and formalization are feasible

What is the direction in the paper? What are the problems in the fields involved in the papers, and the difficulties in the author's research continue

What is the pattern of the pattern, what is the point of view of the paper, and this.

It is very helpful to link reading and programming. If you are interested in a field, reading some arguments

After the text, try the "Toy" version of the procedure described in the paper. This will undoubtedly deepen understanding.

Sadly, many AI laboratory is born to be unused, and the members inside are mainly read and quote their own school laboratories.

jobs. You must know that other institutions have different ways of thinking, it is worth reading, treating seriously, and references

Their work, even if you think you know their mistakes.

Someone will always give you a book or a paper and tell you to read because there is a very flash place and / or

Can be applied to your research work. But waiting for you to read, you find that there is nothing special flash, just just

Strong can be used. So, confusion is coming, "Why don't I?" I missed what? ". In fact, this is because

When reading a book or the paper for your friends, some ideas have been formed in the mind, seemed to you

Research topics have valuable places.

3. Establish relationship

One or two years later, there have been some ideas for the subsequent areas of themselves. At this point - or earlier - join

Secret Paper Passing Network is very important. This informal organization is what artificial intelligence is really doing

Reflection. The work of booting the trend will eventually become a formal published papers, but at least after the cattle fully understands a year.

That is to say, the cow is at least one year for the new ideas.

How does the cattle find new ideas? It may be listened to a meeting, but it is most likely to come from Secret Paper Passi.

Ng network. Below is a general situation of this network work. Jo Cool has a good idea. She will still be incomplete

Nowadays, other work is integrated, write a draft paper. She wants to know how this idea is, so

She sends the copy of the paper to ten friends and ask them to comment. Friends think this idea is great, but also

Out of the mistakes, then these friends copied the paper to their respective friends, so continued. a few

After a month, JO has made a lot of revisions and sent to AAAI. After six months, the paper is officially issued in five pages.

Table (this is the space allowed by the AAAI Conference). Finally, Jo started to organize the relevant program and wrote a longer papers.

(Based on the feedback obtained in AAAI). Then send it to the AI ​​journal. The AI ​​journal will spend about two years.

For the evaluation of the paper, including the time spent on the paper and the amendment, and the corresponding publishing delay. Therefore, in the ideal case, JO's idea finally published in journals for approximately three years. So the cattle rarely can publish journals from this field

What is going on in the article, it's too late.

You can also become a cattle. Below is some of the lyrics of establishing an academic relationship network:

Have a lot of email lists that discuss a AI sub-domain (such as connectionism or vision), select the list of interest yourself

Joining.

• When discussing your own ideas, they are likely to evaluate your thoughts directly, but said:

"Do you read a certain?" This is not an asking, but it is recommended that you read a copy of the literature, it is likely to think with you.

The law has a relationship. If you haven't read this document yet, get detailed information from this document from the master talking to you, or

I will borrow a copy from him.

? When you read a copy of the paper that makes you feel excited, copy five other five people who are interested in it. they

It may feed back good advice.

• There are many discussion groups for informal (sustainable development) papers for different subsequent fields, they per week or every week.

Two weeks ago, discuss all the paper read.

Some people don't mind others to go to the desk, that is, go to read them soon they are not happy on the desk.

Read or often read the paper. You can go over and have you interested. Of course, you must first get the Lord.

People's license, we must know that some people really disgust someone to turn their things. Try those who are close to the people.

Like, some people don't mind if you look down on their files. In the laboratory, there are many people who have learned a deep learning.

There are also a lot of baby in the file cabinet. This is usually faster and more reliable to find papers as compared with the school library.

The way.

• As long as you write something yourself, copy the copy of the draft to those who may be interested. (This also has a potential

Problem: Although there is very little plagiarism in the AI ​​field, it is true. You can write on the first page "Please don't photocopy or lead

Use "to make some precautions.) Most people do not read most of their thesis they receive, so if only

A small number of people returned to you, nor too much. You can repeated several times - this is necessary for journal panels.

Note that except for your mentor, it is generally rarely to the same manuscript.

• After you write a papers, give the copies of the paper to those who may be interested. Don't think that people naturally

Will read the journals or conferencing published by published papers. If it is an internal publication (memo and technical report)

Not easy to read.

The more people you keep in touch, the better the effect. Try to differ from different research groups, AI laboratory, different academics

Switching papers in the field. Make yourself a bridge that exchanges the two research groups that have not contacted, so fast, your

A large-related papers will appear on the table.

If a paper quotes some things you are interested in, do your notes. Maintain a reference for yourself

Log. Go to the library to see if you can find these papers. If you want to know the development track of a topic, you can have

It is a reference to "Reference". The so-called reference diagram is a network of guidelines: thesis A reference

B and C, B reference C and D, C reference D, and so on. Note those papers that are often cited, which is usually worth reading. Participate

The drawings have a wonderful nature. One is often the study group that studies the same topic does not understand each other. You search for the picture,

Suddenly discovered into another part, this usually appears in different schools or different methods. As well as

It is very valuable to understand a variety of methods, which is better than a very deep understanding of a certain approach.

• Temporary shelves. Talk to others. Tell them what you are doing, and ask what people do. (If you are shy to discuss your own ideas with other students, you must insist on talking, even if you don't have any ideas, discuss them with them.

I think there is a good paper. This will naturally guide what to do next. ) Everyday noon at the activity building

The seven floors have an informal lunch seminar. In our laboratory, people are accustomed to work at night, so when lunch

You can discuss with others to make a loose group.

? If you communicate with the outside world - do demonstrate or participate in the meeting - to the discipline of business cards, mainly

The name is easy to remember.

From a certain time, you will start to participate in the academic conference. If you really participate, you will find a fact,

Almost all conference papers are gathered or stupid. (This reason is very interesting, but it has nothing to do with this article.

Do not discuss it). Do you still go to the meeting? Mainly in order to get people outside the laboratory. Outside people spread

Reporting your work, inviting you to report, tell you the characteristics of the academic atmosphere and researchers in a certain place, introduce you

To others, help you find a summer work, such as this. How to get to someone else? If you think someone

The text is valuable, running, said: "I appreciate your paper" and ask questions.

• Get the opportunity to make summer work in other laboratories. This way you will meet another group, perhaps you will learn another

A way to look at things. You can ask how the high grade classmates get this opportunity, they may have already wanted to go

Where you have worked, you can contact you.

4. Learn other fields

Usually, you can only do things in the AI ​​field, don't know about things outside the AI ​​field, as if some people are now

Still thinking so. However, now requires good researchers to understand a few related areas. Feasibility

It does not have sufficient constraints for what is intelligent, and other forms of constraints are given, such as psychology.

Learn experience data. More importantly, other research fields give you new tools you think and look at intelligent new party.

law. Another reason for learning other fields is that AI itself does not evaluate the standard of research value, all borrowed from other

field. Mathematics will be made as progress; the project will ask if an object is reliable; psychology requires repeatable test

Philosophy has strict thinking; etc. All these standards are sometimes played in AI, familiar with these standards help you

Evaluate the work of others, go deep into their work and protect their work.

After six years or so, you can get MIT's PHD, you can lay a solid foundation in one or two non-AI fields.

There is a reading level in more fields and must have a certain degree of understanding of most content. How is the following?

Learn some ways to learn very little field:

? Elective a graduate course, which is very secure, but it is usually not the most effective way.

? Read the textbook. This method is also good, but the knowledge of the textbook is often outdated, and it is generally highly proportional and inner.

Unrelated rhetoric.

? Find out what is the best journal in this field, ask the high people in the field. Then I find out the text worth reading in recent years.

Chapter, and track related references. This is the fastest way to feel the field, but sometimes you may have a wrong reason.

solution.

• Find out the most famous scholars in this field, read the books they have.

• Bubble with graduate students in this field.

• Refer to the foreign school studying the course table in this field. Visiting the institute in the institute, selecting a useful literature.

Here are some subjects related to AI:

Computer Science is the technology we use. The primary graduate courses you need to be repaired must not let you keep your computer.

Science has enough understanding, so you have to learn more knowledge through reading. All areas of computer science - theoretical architecture, system, language, etc. - must be learned.

? Mathematics may be the most important discipline that needs to be understood. For those who work in vision or robots

The essential. For the system-centric work, on the surface, it is not relevant, but mathematics will teach you useful thinking.

You need to read theorem, if you have the ability to prove theorem, you will have a deep impression in most people in the art.

Few people can self-study mathematics, light to be a listener is not enough, but also have to do the topic. Elective as much as possible as soon as possible

Mathematics classes, after the course in other fields is also easy.

Computer science is based on discrete mathematics: algebra, chart, and so on. If you want to work in reasoning,

Logic is very important. Logic is used lot in mit, but in Stanford and elsewhere, this is the mainstream of thinking

method. So you have to have enough logic knowledge so you can protect your own point of view. Opposite MIT Mathematics

Two courses are enough. If you study interest in perception and robot, then not only discrete mathematics, but also need continuous

mathematics. In terms of analysis, differential geometry and topology have solid foundation will provide you with the most common skills. Statistics and

The probability is only usually useful.

Cognitive psychology and AI share almost identical views, but practitioners do have different goals, they mainly

It is an experiment instead of writing a procedure. Everyone needs to know some knowledge of cognitive psychology. In Mit, Molly Pot

Ter opened a good primary graduate course about cognitive psychology.

• If you want to do the job about learning, then developing psychology is very important. Develop psychology from the general sense

It is also very useful, it can tell you what is easy for human intelligence. It also gives the relevant recognition

Cognitive model for knowing the architecture. For example, the work about children's language acquisition has applied solid on language processing theory.

constraint. In Mit, Susan Carey has opened a very good primary postgraduate course for developing psychology.

What is more "soft" in psychology, such as psychological analysis and social psychology, the influence of Ai seems very small, but has

Potential significance. They will give you a very different understanding of people. I am like sociology and anthropology

Social science can play a similar role. It is useful to have a variety of views. You need to learn from the above discipline. Unfortunately

It is difficult to distinguish between these areas which is garbage. Go to Harvard to learn: For MIT's students

It is easy to cross the course of Harvard.

? Neuroscience tells us about the human body to calculate hardware. With recent neuroscience and cough

Rise, it has a very much impact on AI. Mit's Brain and Behavioral Science Department provides a very good course, visual (Hildr

Eth, Poggio, Richards, Ullman, Mobile Control (Hollerbach, Bizzi) and Ordinary Neuroscience (

9.015, grouped by the expert group).

• If you want to study natural language processing, linguistics is very important. Not only that, but it also contains a lot of people related to humans.

Cognitive constraints. In mit, linguistics is mainly responsible by the Chomsky College. You can see if it is in line with your own

interest. George Lakoff recently published book "Women, Fire, And Dangerous Things" can be used as another one

Example of research procedures.

Project, especially motor engineering, has been used as a research area as a research area. Our laboratory is cultivated

There are many requirements that need to do something, such as analysis circuits. Understanding EE also helps to build customization

Chip or debug your own power on your LISP machine.

Physics has a powerful impact on those interested in perception and robots. ? Philosophy is a frame that is invisible in all AI fields. Many Ai work has a philosophical impact. Learning philosophy can help

Help you use or read the views used in many AI papers. Philosophy can be decomposed along at least two orthogonal axes. Philosophical

Often a philosophy of something; philosophy of thinking and language is more relevant to AI. Then there is a variety of philosophical schools, from

Amazing range, philosophy can be divided into analytical philosophy and mainland philosophy. Analysis of philosophy about thinking views and ai field

Most researchers are consistent. Continental philosophy has a very different way of viewing many things happened often. It has

It is impossible to prove AI by DreyFUS. Not long ago, several researchers believe that the Chinese philosophy and AI is compatible.

Different methods of solving problems. MIT's philosophy is an analytical philosophy, and the Philosophy of Philosophy is deeply received by CHO.

MSKY effects in linguistic work.

It seems that there is too much thing to learn, is it? That's it. Be careful with a trap: thinking about all X, "only

There is more to know about X, this problem will become easy. "To know, something related to it needs to know more

It is never finished, but in the end, you still have to sit down and solve the problem.

5. Notes

Many scientists have the habit of making scientific research notes, you should also do this. Maybe you have been told from the fifth grade, for

Every science course should be told, it is true. Different nonsense methods apply to different people, you can do online

Notes, record on the notebook or note. It may be necessary to have one in the laboratory, and there is one at home.

Record your own ideas on the notebook. Only you will read it yourself, so you can remember more casual. record

Your own thinking, problems encountered in current work, possible solutions. Small references that may be used in the future

.

Turn on your own notebook regularly. Some people will be a monthly summary to facilitate future references.

Things in the records in the notes can often be used as the backbone of a paper. This will make life easier. Instead, you will

Found a rough paper - title, summary, subtitle, and body fragment - is a record of your current work

Effective ways, even if you are not preparing to turn it into a real papers. (You may change you for a while

law).

You may find that the book "Notebooks of the Mind" is useful, this book is not

Describe how to do a note, it describes how innovative ideas appear with the accumulation of thought pieces.

6. Writing

There are many reasons for writing.

? In the process of the entire study, you need to write one to two (this depends on the provisions of your department) graduation thesis,

Get PHD or MS.

? Diligent writing is not only the opportunity to practice.

• Academic rules are either published or rot. In many areas and schools, this usually begins to become a teaching

Time, but many graduate students in our laboratory have begun to publish their paper. Encourage publication and distribution

Wen is a good policy.

? Write down your own idea is a good way to adjust your ideas. You will often find self-righteous ideas once

It is said that it is ignorant.

• If your work is to serve yourself, you must publish it. This is also the basics of research

responsibility. If you write wonderful, there will be more people to know your work.

? Ai but it is difficult to do with a single game, you need to get feedback from others frequently. Comment on your paper

It is the most important form. Everything is going to do it.

? Read books on how to write. How to Basic Should Strunk and WHITE

Shouldn't be a introduction. Claire's "Houghton Mifflin is about how to edit in the sentence level. Jacques Barzun's "Simple and Direct: a r Hetoric Fo

R writers "(Harper and Row, 1985) is related to how to make.

• When writing the paper, read the books written in a superb book and think about the author's syntax. You will find unknowingly, you

Already absorbed the style of the author.

? To become a master, it is necessary to pay a lot. After a few years, you must endure and seriously treat others. except

In addition, there is no shortcut to go.

? Writing is sometimes very painful, it seems to be distracted from "actual" work. But if you have already mastered

Writing skills, it will be very fast. And if you regard writing as an art, you can get a lot of fun from it.

interest.

? You will definitely encounter the way of thinking, this has a lot possible reason, and there is no way to avoid it. pursue

Perfect may cause ideas to block: no matter what you start, I always feel that it is not good enough. To understand writing is a process of debugging

. Write a draft first and return to the revision. Write a draft contributing to the way to think about it, if you can't write a text, then write

Outline. Differentially refine until it is easy to write the contents of the child portion. If the draft is also written, hide

All windows that are writing, then enter what you think in your own head, even if it looks like garbage.

When you have written a lot of text, reopen the window and edit the things you just wrote.

Another error is to read all the content sequentially. Usually you should write the core content of the paper,

Finally, it is an introduction. Another reason that caused the author's idea is unrealistic to think that writing is very easy.

situation. Writing is time consuming, if you find yourself can only write a page every day, don't give up.

Perfectionism may result in the paper that is already good enough to be polished. This is a waste of time. (It also

It is a manifestation that has unintentionally escaping to do research). Watch the paper as a sentence when talking to other people in the field

. In the conversation, not every sentence is perfect. Few people will look forward to your own conversation is all

It is a last exchange with the other party.

? Write a letter is a very good practice. Many technical papers, if its style is more similar to the letter to a friend, then there will be very

Large improvement. Adhere to the diary is also the method of practice writing (will also make you experiment with more style, not just technical theory.

Text). These two methods have other substantive effects.

A common trap is spent for a lot of time to pursue rhetorical instead of content. To avoid this. Latex is not perfect, but

It is there much a lot of modifications you need. If this is not enough, you can borrow some people from other people engaged in this research.

Word usage. Many sites (such as MIT) maintain a library of writing rhetorical.

? Clear what you want to express. This is the most important factor in clear writing. If you wrote a bad thing,

And don't know how to modify it, this is likely because you don't know what you want. Once you know, you have to say it.

, Let's say it.

The writing of the paper should help readers find what you do. Whether it is a paragraph organized or artistic organization,

To place the core part in front. Writing a summary. Make sure the abstract has reflected your good idea.

Make sure you understand what your innovation is, then express it in a few words. Too many papers are only general

Introduce the paper, saying that there is a good idea, but it doesn't say anything.

Don't use big words to sell your work. Your readers are very good people, integrity and self-esteem. Contrary to it, nor

Be apologize for your work or decrease.

Sometimes you realize a clause, sentence or paragraph is not good enough, but I don't know how to modify it. This is because you can't get it in the dead alley. You need to return to rewrite this part. This situation in reality rarely occurs.

? Ensure that there is a central idea in your own paper. If your program solves the problem X in 10 milliseconds, tell the reader, you are like

What to do. Don't just explain how the system is constructed, what is it, but also explains its working principle and price?

The value is.

? Writing is given to people, not machines. Therefore, it is easy to understand that the light view is correct. Don't rely on readers yourself

To reasonin, unless it is the most obvious inflation. If you explain the working principle of a little bit on the footnote on the seventh page,

Then there is no further explanation in the twenty-third page, it will be quite referred to if the reader feels confused.

. The formal papers should be difficult to write. Do not imitate the literature in mathematics, their standards are as few as possible

Release, the more difficult the reader feels more and better. This does not apply to AI.

After writing a paper, delete the first paragraph or a few words. You will find that it is a general language that is unrelated to the content.

A better introduction statement begins at the beginning of the first paragraph or the second paragraph.

If you wait until all work, you will lose a lot. Once a scientific research project is started, we must develop

Such habits: Writing explains the current working progress or informal papers to learn from each month. From your research pen

The records start. Write down for two days - If you spend longer, you are a perfectionism

By. Share the paper with your friends. Writing a draft - not to be referenced. Ten papers copy the paper

For those who are interested in (including your mentor). Compared with the formal papers, this is much better.

Division (comment, clarifying ideas, writing exercises, etc.), and in a sense, there is no need there much. Frequently

If you do good, these informal papers can be used as the backbone content of the formal papers, that is, from the AI ​​experiment.

Working Paper in the room became an journal article.

Once you become a member of Secret Paper Passing Network, there will be many people to give you a copy request comment.

. It is very valuable to get the comment of your own papers. So you have reviewed the papers, you get support

There will be more people to comment on your paper. Not only that, learning to evaluate others' papers also help you choose.

Comment on the paper is an art.

• To write a useful comment, you need to read both the papers. The first time I understand its idea and start comments second.

? If someone has repeatedly made a mistake in the paper, don't mark each time. But what is going to figure out what it is.

Why did he do this, what can you do, then clearly indicate or private communication at the first page.

The author of the paper will follow the principle of minimal modification when merging your comments. If you can, only modify one word

If you don't modify a phrase, you can no longer modify the entire sentence. If some of his papers make him

The whole paragraph, the whole section or even the whole papers must be modified, and the letters of the big font should be used to point out so that he is not

Will neglect.

• Don't write devastating criticisms such as "garbage". This is no help for the author. Take time proposes constructive construction

. It is necessary to think of the author.

Comments have many kinds. There is a comment on expression, there is a comment on the content. The comments of expression can also be very different, it can be

School pair, punctuation, spelling error, word loss, etc. You should learn some standard editing symbols. Can also be calibrated

Grammar, rhetoric, and chaotic paragraph. Usually people will continue to commit the same grammatical error, so it takes time

It is clear that it is clear. Next is a comment on organizational structure: Different degrees (clause, sentence, paragraph, subtraction or even chapter) order chaotic, redundant, independent content, and loss of losses.

It is difficult to describe the characteristics of the comment on the content. You may suggest that the author expands its own ideas, considering a problem, error

, Potential problems, express praise, etc. "Because Y, you should read X" is a common use of a useful comment.

When you are asked to comment on the paper, you first want to figure out which comments are more useful. For early discipline, need

You mainly review the organizational structure of content and thesis; for the final draft, you need the details of your main comment.

Note that as a courtesy, before requiring someone to comment, you should first use the spell inspector to check your paper.

.

You don't have to accept all your opinions, but you must take it seriously. It is quite sad to cut some of the contents of the paper.

But often improve the level of the paper. You often find an opinion that it doesn't point out problems, but solve your feelings.

It is unacceptable, then look for the third road.

Publish more papers, which is actually easy than thinking. Basically, the standard of the AI ​​publication reviewers is: (a

) New intentions; (b) in certain aspects, in line with standards. Look at the IJCai's conference, you will find the standard of the paper's admission.

low. This situation becomes worse due to the inherent randomness of the review process itself. So a word published

It is kept try.

? Ensure that the readability of the paper is better. The cause of the paper is rejected, in addition to meaningless, it is impossible to understand or group

Woven.

• The paper should be exchanged for a period of time before the journal, and appropriate amendments according to feedback. Boycott

The practice of rushing the results in a hurry. In the field of AI, there is no competition, and no matter what to say, the publishing cycle

The delay is much more than the time to comment on draft.

Read a journal or conference you want to submit to ensure that the style and content of your paper are suitable.

? Many publications have a "authors of the authors" and take a closer look.

• The main conference will be awarded the award-winning papers that have been accompanied by the content and expressing the awards in the received papers.

? Usually, it is a relatively short report of some work content to the meeting, and then processes the journal.

A final formal papers of a length.

The papers have been decided - don't blaely.

The paper review process of the journal and the conference is very different. In order to save time, the review of conference papers must be rapid

There is no time to abolish or communicate. If you are rejected, you have failed. But the journal papers are different, you can

Temporary and editorial arguments, by editing and reviewers arguing.

• Reviewers generally help you. If you have received a more annoying review report, you should to the General Assembly

Chair or editorial complaint. You can't expect how much feedback can be obtained from the report of the conference paper review. Direct

The paper is often a great suggestion. You don't have to do full according to the recommendations of the review report, however, if

You don't do it in accordance with the report, then you must explain the reasons, and it is necessary to realize that this may result in further negative evaluation.

. Anyway, no matter which review, it is polite as a reviewer. Because in the rest of your career,

You will be with the accused in an academic circle.

? MIT AI LAB MEMOS is generally or close to the level. In fact, Technical Reports basically

Is a revision of these MEMOS. Working Papers is not official, this is a good point to share your own papers.

Methods of things. To publish these internal files, just receive a form to the publications office (eight floors), and two faculty signed.

Just like other research activities, the time spent on the paper is always higher than expected. The published of the paper is time consuming

This issue is more serious. When you have completed a papers, cast it out, wait for the publication. After a few months, the paper

Theory is returned. You have to modify the paper. Then it is a few months before returning to your modification confirmation. Such as

If you express the different forms of the paper, if there is a short investment, a long period of time, such a process

Several rounds will be repeated. As a result, it is possible that when you are tired, the research theme has also been tired of years, you still

The paper is modified. This revealed that we don't want to do those who need enthusiastic investment but it's hard to publish papers -

- Hard sense.

7. Speech

Another way to communicate with peers is the speech, the problem of writing the paper mentioned above, the same applies to speaking

. Standing in front of the audience, it is calm and will not make the audience's ability to sleep. For you to successfully recognize others,

Dear and even the final job hunt is critical. The ability of the lecture is not born, and below is some learning and practice speakers.

Methods:

Patrick Winston has a very good text about how to do a speech. Every year, he will tell this.

Perform, demonstrate and describe its speech skills.

? If you feel that you are a bad speaker, or you want to be an excellent speaker, choose a public speech.

class. The primary performance class is also useful.

• If your mentor has a regular research semina, voluntarily gave a speech.

• Mit Ai Lab has a range of semicona symposium called Revolving seminar. If you feel some of yourself

The view is worthy of writing into the AI ​​Memo or the discussion of the discussion, and the self-sufficiency is brave to make a report.

In-depth understanding of different robots in the laboratory, when you come to friends, you can take them around

A circle and do a 60-minute report on the robot.

Some people think this is a good way to find thoughts is far more than modified papers.

(Nike Brady once again said that all of his best papers came from a speech).

? Exercise in an empty room, the best is the report you want to do immediately. This helps adjust the skills of the report: every one

What is the slide saying; the delay of the conversion and keep it smooth; maintain the synchronization of the interpretation and slide; the length of the estimated report

. The less time you spend on the adjustment device, the longer the time you stay with people.

• Exercise with mirrors, recorders or recorders is another method. The laboratory has these three devices. This also helps to adjust

Both your own pronunciation and body language.

• For comparative official reports - especially your reply - should practice it in front of a few friends, please batch

Evaluation.

• Observe how others are reported. There are a lot of people who visit MIT will be reported. Reports to participate in this

Unfamiliar areas, and if the report is unresponsible, you can secretly analyze where the speaker is wrong.

? Find a friend, tell him the nearest idea. This can improve communication skills, and debug your own thinking

road.

8. Program design

Not all AI papers contain code, and many heavyweight characters in the field have never written an important

program. But in order to initially approximate AI work principles, you must design it. Not only many AI research work

To write code, and learn how to design can give you what can be calculated, what is unfoundable, this is ai

Cognitive scientific contributions.

In MIT, all AI programming in nature uses Common Lisp. If you still don't know, hurry to learn. of course,

Learning a language is not equivalent to the learning program design; some of the technologies included in the AI ​​program design are different from those used in system programming or application design. When you start learning, you can take a look at Abelson and Sussman.

"Structure and Interpretation of Computer Program" and do some exercises. This book

The AI ​​program is not coheed in nature, but contains some of the same technologies. Then read Winston and Horn written LIS

P books third edition, there are many elegant AI programs in books. Finally, perform practical programming, not reading, is the most

A good learning process.

Learning LISP programming has many traditions. Some people are used to writing code, depending on the personality. Others looking for machine

Will learn from experienced programmers or ask him to evaluate your code. Reading the code of others is also very

Effective method. If you can go to the senior classmate to ask for their source code. They may have some complaints and say their own programming

The style is poor, and the program does not work cloud cloud. Anyway, in the end you got the source code. Then you want

Read more carefully, this takes time. Usually read and fully understand the time spending time with others and your own programming

The time is more than the time, so you have to plan a few weeks of one or two semesters in your head.

People's code. You will learn from the middle school, I have never thought of tips in the textbook. If you read the big paragraph

Segment can not understand the procedure without comments, you will understand how to write code.

Those knowledge learned in the software engineering class is still useful in the AI ​​programming. To add a comment to the code. Use correct

Data abstraction. Align the map and your code, because the language you use is basically Common Lisp, therefore

The plantability is very good. Such this.

After several years of learning, you should write some of your own standard AI modules, such as:

? True value maintenance system

? Planner

? Rule system

? Different styles of interpreter

? Optimized compiler with process analysis

? Frame system with inheritance characteristics

Several search methods

Explanation-based learning

Anything you interested can try to implement. You can seize the spirit of the problem, complete one within a few days

Feature versions. Modifying existing procedures is another effective way, provided that you have already written like this, and

And doing problems such as working principles, advantages and disadvantages, and efficiency.

Unlike other usual programmers, there is little mutual borrowing code between AI programmers. (Demo code exception). This part

There are very few truly roles in the AI ​​program. (Many famous AI procedures are only in the three examples mentioned by the author thesis

It works, although this situation has been improved recently). Another reason is that the AI ​​program is usually in a hurry, and

There is no problem that the generalization is considered. Using FOOBAR's "Standard" rule interpreter, it is very effective at the beginning, and will soon send it.

It is now lacking some features you need, or not enough efficiency. Although you can modify your own needs,

But remember to understand the code of others is very time consuming, sometimes it is better to write one. Sometimes build a standard package

The work itself can become a paper.

Like the paper, the program may be too pursuit of perfection. Do not stop overwrite the code to make perfect, maximize abstract

Things, write macro libraries, dealing with the operating system core, which makes many people from their own papers, deviation

The area of ​​your own. (From another aspect, this may be the means you need to make a living).

9. Tutor

In MIT, there are two types of mentor, teaching tutors and the paper tutors.

The work of teaching mentors is relatively simple. Every graduate student is allocated a teacher in the department as a teaching tutor. teaching

The role of the tutor is the representative of the system, telling you what is the formal request for you, if your progress is slow, urge you, ratify your course plan. If everything goes well, you only need to see the teaching tutor twice a year, in the registration day

. On the other hand, if you have trouble, the teaching tutor reflects or guidance for you to the system.

The paper tutor is a person who supervises you. Choosing the paper tutor is the most important choice during your study, it is important than the topic.

many. In a broader sense, AI is learning through the master of the apprentice. There are many technical aspects of the field or

Informal knowledge in the research process can only be learned from the instructor, can't be found in any textbook.

Many AI majors are a weird person, and the graduates are the same. The relationship between the mentor and the graduate students is very personal, you

Personal characteristics must be well matched with the mentor so you can work together.

Different mentors have different styles. Here are some factors that need to be considered:

What exact guidance do you need? Some tutors will give you a well-defined problem that is suitable for the papers.

The law is explained and tells you how to work. If you are in a place, they will tell you how to carry out

go with. Other tutors belong to a hand, they may have no help to your topic, but once you choose the topic, they

It has a very big role in guiding your ideas. You need to consider clearing yourself for independence or need guidance.

How much does you need? Some mentors have asked to meet you every week and listen to the report of your progress. they will

Tell you the paper you should read and give you actual exercises and projects. Other tutors will not exceed your semester

Twice.

How big is the pressure you can withstand? Some of the pressure applied to some mentor is very large.

How much emotional support?

? What is the serious level of listening to mentor? Most tutors will quite officially recommend your paper topic. Some tutors

It is trustworthy, the recommendations they give, if implemented, almost certainly will make an acceptable papers

If it is not an exciting papers. Others throw a lot of ideas at once, most of them are unrealistic.

But there are some, perhaps leading to major breakthroughs. If you choose such a tutor, you must first take yourself as a

filter.

? What type of research group is a tutor? Some professors will create the environment and gather all the students together, even

They are not doing the same project. Many professors meet with their students every week or every two weeks. Is this useful for you?

? Can you get along with the professor's students? Some students have found them to build good with students in other teaching groups.

Working relations.

Do you want to participate in a big project? Some professors will decompose a large system, and each student is part of it. This gives you with a group

The opportunity to discuss problems.

Do you want to be supervised together? Some paper projects contain multiple AI fields, need you to build secrets with more than two professors

Sliced ​​work relationship. Although your formal papers mentor has only one, sometimes this does not reflect the actual situation.

? Does the mentor are willing to guide the topic of the paper outside the research field? Can you work with your mentor, what to do?

Body is more important. MIT's robotic is guided by quantum physics and cognitive modeling papers;

Guide the visual papers. But some teachers will only guide their own fields in the field of interest. This for that

This is especially true for young teachers who want to get lifelong positions.

? Do you have a struggle for you? Some tutors will fight for you or some hostile entities

. Sometimes the system is unfavorable to certain types of students (especially for female students and quirks), this is very

important.

? The mentor is willing to recommend your work at the meeting? This is part of the mentor work, working for you in the future

Righteousness.

These factors described above, different schools are very different. MIT provides much free freedom compared to most schools. Looking for a paper tutor is the most important task of your graduate student's first grade. At the end of the study, or the beginning of the second school year, you must

There must be a paper tutor. Here are some kind of trick:

• Check the research summary of the laboratory. Among them, there is a parameter that describes each teacher and many graduate students.

do what.

• If you are interested in the research work of some teachers, check it out.

? In the first semester, talk to as many teachers as possible. Feel your favorite, their research and guidance style

What is it.

? Talk to the graduate students expected. To ensure multiple students with the mentor, because each mentor is in different studies

There are different working methods and communication effects when conducting. Can't be seen by a student's view.

? Many teachers' conferences in the study group are open to new students. This is a very good way to understand the way the instructor works.

path.

As a discipline, Ai is a very common point that many useful work is done by graduate students, not doctors -

They are busy doing management. This has several consequences. First, whether a teacher's reputation will be hired for life, it is very

Depending largely, depending on the student's work. This means that professors have a strong motivation to attract the best students to work for themselves.

And give efficient guidance and sufficient support. Another consequence is that because most students' papers are by tutors

Form, therefore the direction and development of the entire field depends largely on what kind of graduate students choose to choose.

When the tutor is selected, determine the requirements of your tutor, make sure that the mentor knows. Don't waste because of communication, waste

Time is on the project you don't want to do.

Don't completely depend on your mentor, to build your own network. Find some people who can regularly review your work is very important

Because it is easy to walk into the magic because of research. People in the network can include graduate students and old in their own laboratory or external units.

division.

In the relationship with other students, teachers and even their own mentors, it is likely to encounter racists, gender discrimination, the same

Fat or other awkward things. If you missed, go seeking help. MIT's ODSA published a call

Do a lot of suggestions in the booklet of "Stop Harrassment". "Computer Science Women's Rep

Ort ", can be found in the LCS document room, also related to it.

Some students in the laboratory are only guided by tutors in nominals. This is very good for those who have strong independence. But if

If you have completed the work of a mentor, unless you don't have a tutor, you have a firm support network.

Otherwise, don't do it.

10. Paper 10.

It is mainly to do research, including topics, including topics, including topics, including topics.

More time consumption.

The purpose of the master's thesis is to make a doctoral paper training. If you are not ready, you are not ready.

of. The most nature of master's thesis is to show your master: You have fully understood the latest developments in the field,

And have the appropriate level of operation. Don't need to expand the latest knowledge in this field, nor does it require your discussion.

Text. However, our laboratory papers is always atmospheric, so many master's thesis actually develop in the field.

Significant contributions have been published. This is not necessarily a good thing. Many people have concentrated in the company

The work of the taxi, so Mit has a reputation: the quality of master's thesis is often high than doctoral thesis. This is contrary to Master

Originally, the original purpose for Ph.D. Another factor is that the research must contribute to the field, at least

It takes two years, which makes the graduate study time difficult to endure. Maybe you can't feel in a hurry, but when you

After seven years in the laboratory, you must not wait to escape. Master's graduation average time is two

Senior, however, the computer system strongly encourage students to graduate in advance. If a master's topic is too large, you can decompose, some to do a master's thesis, another part to doctoral thesis.

Want to understand what is the study of master's thesis, read a few newest master's thesis. Remember that the better papers are those

The version or becomes a technical report, because this marks that the paper is considered to be the latest knowledge of the field -

In other words, their papers beyond the level of master's thesis. Also read some papers through but no publishing,

All pass the paper can be found in the MIT Library. Doctoral thesis must have expanded the latest knowledge, doctoral thesis

Research must have the quality of published. MIT's temperament has been manifested, and many doctoral thesis have been in a few years.

Authority in a sub-field. For MIT's doctoral thesis, create a new field, or propose and solve one

New problems, nothing is nothing. Although this is not required.

In general, it takes two to three years to do doctoral thesis. Many people spend more than two years to say to master's degree.

See, and topic. During this time, you can try some other things, such as making assistants or lay down in a non-AI field.

Solid foundation or organize a band. The actual writing time of doctoral thesis is about one year.

The topic is the most important and most difficult part in the work of the paper:

Good paper topic not only express personal views, but also communicate with peers.

• Select topics must be you willing to pour enthusiasm. Personal vision is the reason you as a scientist, it is you

The most concerned image, principles, ideas or goals. There are a variety of forms. Maybe you want to create a calculation that can be conversantly

Machine, maybe you want to save human beings from the stupid use of the computer, maybe you want to show everything is unified, or

Let you want to find new lives in space. Vision is always big, your paper does not realize your vision, but

Can work in that direction.

When doing papers, the most difficult thing is how to reduce the problem to a resolved level, while the scale is enough to be a paper

. "Solving the width of AI" is an example of common problems, and the topic is too great. You will find that it needs to be constant

Reduce the range of topics. The topic is a gradual process, not a discrete event, will continue to you to announce the paper

After that moment. In fact, solving problems is usually much easier than accurately describing problems. If your goal

Is a 50-year project, what is the reasonable ten-year project, one year? If the target is huge, then

What is the most core component, how to maximize the core components?

? An important factor is how much the risk can be tolerated. Weighing between the final success and the risk. This

It is not always right, many researchers have not been involved in the AI.

? Good papers have a central part, you are sure that you can do it, and you and your mentor agree this already

Meet the graduation requirements. In addition, there are many expansions in the paper, there is a possibility of failure, but if it is successful, it will increase

The extent of the paper. Although not every paper selection is compliant, it is worth a try.

• Some people think that work in multiple projects can be done when choosing questions. This is indeed reduced

Risk. Other people are willing to choose a separate question before doing any work.

? May only be interested in a certain field so that your topic is narrow. Sometimes, you will find the system

The teacher can guide you to choose from. Maybe it will also find that there is no natural selection in that field.

Instead, there is a good idea to other fields.

? Master's topic is more difficult than doctoral topics, because master's thesis must do not have enough self-confidence when you know.

What is the factor for the doctoral selection needs to consider whether the field of the graduation of the master's degree may expand or be a base

Basis, or simply go to another area. Things to be in the same area is simple, it may take only one to two years, especially if the topic for doing doctoral thesis has been discovered in the work of the master's phase. Inadequate

It can increase the width of the knowledge in the field of easy-to-set shape.

? Some papers are very new, and some are very ordinary. The former has created a new field and explored the currently not studied before.

Elephant, or for difficult problems, the latter is perfectly solved with a good problem.

. Both papers are valuable. Which papers are selected depend on the personal style.

The "future work" part of the paper is a good source of papers.

No matter what topic, you must have never done it. Even if someone works at the same time, it is not good. Have

Many things can be made without competition at all. There is also a common situation, I feel very panic after reading other people's thesis, good

As it has solved your problem. This usually occurs during determining the problem of the paper. In fact, it is often just a surface class.

It is like, so give the paper to a high person who knows your work, see what he said.

• Mit Ai Laboratory is not all relevant to artificial intelligence; some are related to hardware or programming languages,

also.

After choosing a good question, even if you have a little virtual, you have to answer the following questions: What is the argument of the paper? What do you want to clarify?

You have to have a saying, a five-minute answer. If you don't know what you are doing, others will not treat seriously

Your topic, worse, you will be in the topic - the circle of the topic, you can't extricate themselves.

After starting to study the research, you must be able to interpret each part of the theory and implementation of each part with a simple language.

.

Remember, once the topic is selected, you must have a clear consistency with the standard of the instructor to complete the paper. If you and him

The paper has different expectations, and finally you will definitely die very badly. It must be defined as a "completed test", like a series

Can prove your theory and procedure examples. This is a must do, that is, your mentor does not require this. If the ring

The fundamental changes have changed, and the test has to change.

First try the simplified version of the issue. Use an example test. Before formation theory, you have to have a complete exploration.

Phenological example.

There are many ways to waste time during the process of doing the paper. To avoid the following activities (unless it is indeed related to the paper):

Design of expression; excessive payment on user interfaces or graphics interfaces; invented new formal methods; excessively optimized code

Create a tool; bureaucrat style. Any work that is not very relevant to your paper should be reduced as possible.

A well-known phenomenon "Thesis Escape" is that you suddenly discovered that the bug that corrected a operating system is very attractive.

It is also very important to work. At this point you always consciously unconsciously deviate from the work of the paper. Remember what you should do?

. (This article is a paper escape from the paper for some authors).

11. Research methodology

[This section is relatively small, please add]

Research methodology defines what is research activities, how to carry out research, how to measure research, and what is called

success. AI's research methodology is a hodgepodi. Different methods are defined different research schools.

The method is the tool. Use it, don't let them use you. Don't take yourself in the slogan: "AI research needs

Deliberating the foundation, "philosophers will only talk about it, and artificial intelligence needs to work."

What is clear? "In fact, you must be successful in the field of artificial intelligence, you must be good at various technical methods."

It must also have doubtful attitudes. For example, you must be able to prove the aimation, and you must also think about whether the aim is said

What is going on.

Many excellent AI chapters are uniform in several ways to be balanced. For example, you have to choose one in too many gersters.

On (may not be related to any actual problems) and cumbersome implementation (expressing practical solutions expressing 无 无)

The best route. You often face research decisions from "clean" and "dirty". You should take the time to issue problems

Is it in a certain extent? Still maintain the original state of the problem, at this time, although the structure is poor, it is close to the actual situation? Pick

Use the former method (if a feasible) will be clearly determined, but this process is often cumbersome, or

At least it will solve the problem directly. The latter has the risk of caught in the vortex of various processes. Any work, anyone,

Wise balance must be made.

Some work icons science. How do you observe how people learn arithmetic? How is the brain work, how to jump in the kangaroo,

Then figure out the principle to form the theory of test. Some jobs works: strive to create a better problem solver

Or algorithm. Some jobs like mathematics: deal with formal deal, to understand attributes, give prove. Some work is an example

Drive, the goal is to explain the specific phenomenon. The best work is the combination of above.

Method has social, see how others overcome similar problems, ask others how to deal with certain special

The situation is.

12. Emotional factors

Research is a hard work, it is easy to lose interest. An embarrassing truth is student reading in this laboratory

Only very few proportions finally got a degree. Some people leave because they can earn more money in the industry, or

The reason for people; the most important reason is due to the paper. The goal of this section is to explain the cause of this situation, and give

Some beneficial advice.

All studies have risks. If your project is impossible to fail, it is developing, not research. Facing the project failed

How hard is it, it is easy to explain your responsible project failure as your own failure. Although this is actually issued

You have courage to challenge difficulties.

There are very few people in the artificial intelligence area always have been successful, and the paper is out of the past. In fact, failure is often. You can

It is found that they often do several projects at the same time, only some is successful. The final successful project may have repeatedly failed too much

Time. After many after the failure of the method, it took the final success.

In your future career, it will experience a lot of failure. But every failure project represents your work, very

Multi-thinking, thinking, and even prepared code, you find available for another completely different project after several years

. This effect only appears after you have accumulated a considerable failed failure. Therefore, there should be after the initial failure

The belief will work.

The actual time spent on the study is often much more than the plan. A small skill is to assign three times that of each child task

Time (some people add a sentence: "... even considering this principle").

The key to success is to make research becoming part of your daily life. Many breakthroughs and inspiration happen when you walk.

If you think about it, you will find the Spring Festival. Successful AI researcher, persistence

Generally greater than the heterogeneous. "Try" is also very important, which is the ability to distinguish the shallow and important ideas.

You will find that your proportion of success is random. Sometimes, I need three months before I have finished a week.

Work. This is a delightful, making you more willing to work in this field. Every time, you are completely trapped.

There, I feel that I can't do anything. This situation is difficult to handle. You will feel that you will never make any valuable

Things, or feel that you no longer have the quality of the researcher. These feelings are almost definitely wrong. If you

Is a student admitted by Mit, you are absolutely qualified. What you need is to suspend, maintain a high level of bad results

endure.

You have a lot of work to do through a short-term goal in regular settings, such as weekly or monthly. There are two ways to increase the possibility of achieving these goals, you can record the target in the notebook and tell another person. You can be with a

A friend agreed to exchange weekly goals and see who eventually realized their goals. Or tell your mentor.

Sometimes you will completely trap, similar to the way of thinking of writing procedures, there are many possible reasons, but there is no certain

Solution.

The range is too broad, try to solve the child problems in the process.

Sometimes the suspicion of your research ability will throw away all your enthusiasm and make you nothing. Keep in mind research capabilities

It is a skill that is learning, not born.

• If you find yourself in a serious dilemma, there is no progress in a week, trying to work only one hour a day. A few days

After you will find everything back to the right track.

? Fear failure will make research work more difficult. If you can't get a job, ask yourself because

In escaping experiments to test their own ideas. It is found that this time in the last few months is completely, which may

Terviate you further work. There is no way to avoid this, as long as you recognize failures and waste is also a research process.

section.

? Take a look at the book of Alan Lakien "How to Get Control of Your Time and Your Life", including

Incidentally invaluable methods that make you enter the status of creativity.

Many people find their own personal life and research ability. For some people, when life

When the cut is not as good, the job is a refuge. Other people can't work if life is in chaos. If you think

I really have a sad self-extricating, go see a psychologist. A informal survey shows that our laboratory has

Half of students have seen a psychologist during the study period.

One reason why artificial intelligence is so difficult is not universally accepted success standards. In mathematics, if you prove

A certain aim, you have indeed do something; if the aimer does not have an evidence, then your job is a happiness.

. Artificial intelligence has borrowed some standards from related disciplines, as well as some standards. Different practitioners,

Subfields and schools will emphasize different standards. MIT is more emphasized than other schools, but the laboratory is also

There is a big difference. One consequence is that you can't satisfy all people. Another consequence is you

I can't determine if I have made progress, which will make you feel unsafe. The evaluation of your work is from "I have seen the most

Great "to" emptiness, more, unclear, "not enough, this is normal, according to the feedback of others

Your own work.

There are several ways to help overcome the insecurity in the research process. I'm recognized: including the acceptance of the graduation thesis,

Table paper and so on. More importantly, with as many people as possible, you will talk about your ideas and listen to feedback. First, they can contribute

Useful thinking; second, there must be some people who will like your work, which will make you feel good. Due to the progress of evaluation

The standard is so uncertain, and it is easy to blind if it is not fully exchange with other researchers. Special when you don't feel

Too good, you should communicate on your work. It is very important to get feedback and support.

It is easy to see your own contribution, always think: "If I can do it, it must be insignificant. All my thoughts are

Too obvious. "In fact, when you look back, these although it is obvious to you, it is not necessarily obvious to others.

of. Explain your work to many other people, you will find that now is a strange thing for you.

difficult! write down.

A won the winner of Nobel, I have doubt my own problem (in the process of your research, you have always feel that you are

Do you want to shocked the world? The survey indicated that the winners agreed that they often doubt the value and correctness of their work, and they have experienced that their work is irrelevant, too obvious or wrong. Any scientific process

The common and important parts are often strict evaluation, many times, the value of the work is also the scientific process.

Avoidable part.

Some researchers have found that they work better than someone else's work. Although artificial intelligence research is often quite personal

In part, there are some people working together, create a system, and publish papers. Our laboratory already has at least

A union of joint graduation papers. The disadvantage is that it is difficult to distinguish the contribution to the papers. Man with laboratory

Cooperation, such as the summer work, there will be less.

Many students who came to Mit Ai laboratory were the most powerful people in the school. After coming here, you will find a lot.

More smarter. This has formed a blow for many of the students' self-esteem. But there is also a smart person around it.

Benefits: Before you don't have anything you do, but you don't know it), you will be defeated by others.

. More realistic, there are so many smart people in the real world. So find a consultant to find a consultant

Keep psychological balance. First of all, someone will pay for you, this shows that you do something. Second, you find

They really need your help, and work is good to have a sense of satisfaction.

Conversely, every student in the laboratory is selected from more than 400 applicants, so many of our students are very

Big. It is easy to think that only I can solve this problem. This is nothing wrong, and it helps to promote the development of the field.

Potential problems is that you will find that all questions are much more complicated than you imagined, and the time of research is more planned.

More, completely rely on yourself. These make many people in us have fallen into a serious confident crisis. you

It must be facing a fact that you can only contribute to a small part of a sub-field, your paper is impossible.

Solve a major problem. This requires fierce self-evaluation, full of pain, sometimes it takes about a year.

Time can be done. But all this is worthwhile, not self-evident, helps to study with a spirit of a game.

There are at least two emotional reasons for people to endure the pain. One is driven, passion on the problem. You do this study

Because I leave it, I can't live, and many great work is made. Although this also has oil lamp

may. Another reason is a good study is full of fun. In most time, research is painful, but

If the problem is just right, you can play the same solve it, enjoy the entire process. Both are not compatible,

But you need to have a trade-off.

To understand how the study is, how to comfort yourself when you are suspected, read some of the self-songs of contemporary people.

Some effects. Gregory Bateson's Advice To a Young Scientist, Freeman Dyson's Disturb

Ing the universe, Richard Feynmann's Surely You are JOKING, Mr. Feynmann !,

GEO

RGE Hardy's a Mathematician's Apology, and Jim Watson's The Double Helix.

When you have completed a project - such as a papers - After one or two months, you may think that all this is not worth it. This

The submerged effect is due to the long time being suppressed on this issue, and I feel that this can be done better. Always like this, don't

too serious. After another two years, look back, you will feel: Hey, awesome! Have a great job!

Tail

This article contains ideas, text and comments from Phil Agre, Jonathan Amsterdam, Jeff Anton, Alan Bawden, Danny Bobrow, Kaaren Bock, Jennifer Brooks, Rod Brooks, David

Chap

Man, Jim Davis, Bruce Donald, Ken Forbus, ERIC Grimson, Ken Haase, Dan

Huttenl

Ocher, Leslie Kaelbling, Mike Lowry, Patrick Sobalvarro, Jeff Shrager,

Daniel

Weise, and ramin zabih. We would like to thank those who contribute to this article (contributing to our papers.

People, by the way, thank you, especially our tutors.

Some ideas listed above come from John Backus, "On Being A Research" and Alan Bundy,

"How to get a phd in ai" of Ben Du Boulay, Jim Howe, Gordon Plotkin.

转载请注明原文地址:https://www.9cbs.com/read-41933.html

New Post(0)