How to do research during graduate school: some suggestions

This article is translated from "How To do graduate-level research:some Advice". 1. Introduce

This article is intended to provide some useful advice to graduate students, especially Ph. D. Students, and you are about to begin your research career in the Automated Network research group of USC's Electronic Engineering Systems division. Hopefully this article will also provide useful help to graduate students in the same research field as other institutions.
Inexperienced graduate students often have misconceptions about the nature of research. For example, they think that doing research is as similar as class or just need the same skills, and that these research projects are like homework, teachers have ready-made answers and will provide them with some tools. Fortunately, this article helps them to eliminate misunderstandings and to get right into the path of research. Many suggestions may be the same as common sense, but they are rarely taught to students in the classroom because they are common sense.
This is an article on the provision of recommendations, so it is often modified. First, I list a few important aspects in front of each section. And then in the next few months of the course I'll write them as oral reports. 2. Personal and professional principles 2.1 Motives and goals

-Why are you studying this? You should have a clear understanding of this before you read the blog. The reason is not "I can't find a job" or "I don't have anything else to be interested in right now."

-During school, it is important to be aware of goals and motivations in your mind, imagine where you will be and what to do in 5 years, and figure out how to achieve it.

-You must set yourself specific scientific or professional development goals: short term (each semester), medium term (1-2 years), long term (4-5 years). The key is to be specific in the goal. Goals should be quantified: "I think a good study this semester" is not a specific goal. "I want to write a conference paper." "I want to participate in more extracurricular activities" is not a specific goal. "I'm going to participate in X activity this semester."

-During school, you must be passionate and motivated to succeed. The good news is that you don't have to be born with this quality, but you can cultivate and learn it. 2.2 Tracking Progress

-Generate an online resume and web site with research results and update it at all times. Avoid using words such as "this site is being built."

-Your site should disclose your personal research. It can not only advertise your internship (work), but also monitor your professional progress. Try to update your resume every month each semester. 2.3 Time Management

-You have to balance the ability to do a few things.

-Do not prioritize tasks based on the due date and the result and importance. Tasks that help you achieve specific short-and medium-term, long-term goals are the most important tasks.

-Set high priority on important things like paper: Remember that you may be in a mess. Now, you should find someone to tell you when to call homework, when to finish the project and so on.

-Use a schedule to cross out a task every day, which will give you a sense of accomplishment in your daily work.

-set aside some time to plan and summarize at the beginning and end of the day to see where you are and what you are doing on the way to accomplishing your specific goals. 2.4 Get along with people, cooperate and receive guidance

-You will not succeed if you do not complete the task on your own during school or during your professional studies. In order to do better, you need to help others and get help from others. You need to be aware of your interpersonal relationship.

-When confused, follow the golden rule: you have not wanted to others. Be respectful and patient with other people's requests.

-Remember to associate with people is not a 0 and a game, when you win, others do not have to lose what, on the contrary can be a win, that is, to do something that is beneficial to others.

-When it's time to help others in your career, help mentors, classmates and co-workers.

-Ask for help from someone who is more experienced than you, and give advice to those who are seeking assistance from you.

-be cautious about expressing gratitude when you get ideas from others (whether through references or lectures). Do not complain about the shortcomings of others, do not hold on to other people's shortcomings, because it makes no sense and will give you and him to fill obstacles.

-Try to get to know people who work with you, spend as much time with them as possible, and learn about other people in your research area or outside your department, such as through meetings or emails.

-Work with others when possible, and learn to collaborate on research. There are no shortcuts to improving your output. To make an important contribution in the cooperation. 2.5 quality: User Viewpoint

-You have to do the best job possible and keep improving it.

-The fundamental task of making high-quality work is to think from the user's point of view. When you write an article, think about how to make it easier for your readers to understand what you mean and learn from you. Instead of focusing on what you want to say when speaking, focus on what you want your audience to hear or understand.

-When writing articles or making reports for your work, ask for honest feedback and let it provide the quality of your work. 2.6 Attitude

-especially when out of academia, some students start to doubt themselves, using a professional vocabulary is, "imposter syndrome." It is characterized by the inner monologue will say "I do not belong here, I do not have the people around me, I just mixed in, they will soon find that I am just a fake, that I am in trouble." You should be able to adjust the mood and overcome it.

-Try to stay passionate and confident. One of the benefits of setting specific goals and accomplishing some goals each semester is that you build self-confidence each time you complete a task. You need self-confidence to deal with the difficulties of the future in the research process, how occasional papers are rejected.

-Responsible for your professional development. If things don't work out, don't blame them, even if it's someone else's fault. Go straight ahead and think about how to do better.

-Develop initiative. Be proactive. Don't always wait for someone to tell you what to do.

Adhere Don't give up on your task easily. If one method doesn't work, try another method.

-Be interested in your task and believe you can make interesting research. Your skepticism will be reflected in the quality of your work. 2.7 Coping with failure

-Remember that no matter how good you are, you will fail the interview, not once, two times but many times. It may be that your paper has been rejected, the results of the simulation or analysis are not what you expected, and cannot prove the result you want.

-Accept it, learn from it, and figure out how to improve it or correct it, and most importantly, stick to it and try something else if you can't. Reflect on yourself, why HD failed, for example, whether this new area needs other background information. Whether your method is fundamentally wrong. Whether you can avoid the same mistakes in the future. What the judges are saying.

-One of the key points in coping with failure is to do a couple of things at the same time, giving these things a different level of priority. If one thing doesn't go on for some reason, then you can move on to the next thing. Of course you have to figure out how many things you can do at the same time. 2.8 Seize the opportunity: Stay ahead one step

-Always one step ahead of your present situation. As a graduate student, you must be eager to study opportunities. If you're a graduate student, attending a professional event is normal for a person who finishes a Ph. D. thesis.

-for example, peer review, conference speech, a chapter in a book to write a part of a live demo to help your mentor complete the funding application form, attend academic, conference, and seminar activities in the Department, and teach other courses.

-Take the lead role in an organization as much as possible. 3. Research Skills – Basic Skills 3.1 Note

-Use notebooks to write down all the things that are relevant to the research content, in conversations, meetings, discussions, or reading and thinking.

-Often browsing notebooks. 3.2 Read

-Remember that you have to learn how to read the literature, which is different from reading textbooks. One reason is that the literature is not as authoritative as textbooks.

-Any paper, you have to be able to say its contents in a sentence.

-Ask yourself three questions: 1. Is this article up to date? 2. Is this article interesting? 3. Is this article workable? (If other places are defective)

-Learn to read a large amount of literature in a short time. Read the title first, if it's interesting, then read the summary, if it's interesting, reread the intro/results/conclusions; the final reading of the paper is the methodology in this article.

-Read with a critical eye, examine the assumptions made by the author, and think about how it is linked to the conclusion.

-Ask yourself how to expand/improve the job. For example, change the assumptions.

-whether this article is well expressed on the whole.

-only when you read a lot of literature can you begin to appreciate the quality of the article, so start as much literature as possible, and then keep reading a few articles a week.

-Consider reading a large amount of literature in areas of interest to you and writing a review when you begin your research. 3.3 Listen

-Start with as many discussions as possible, whether or not related to your research, as long as it is in the fields of information technology, electronic engineering, and computer science.

-Try to find the point of the discussion.

-Ask yourself questions during the discussion: what is interesting. Which point is new. Whether this method is helpful to yourself.

-Be positive when listening. To ask questions, this is the only shortcut we can learn something.

-Take notes. If you want to know more about the speaker's research, find him and email him. Read the references provided by the speaker.

-Provide feedback to speakers. Especially when you have some ideas about their work or when you know that a related job might be useful to them. 3.4 said

-it is said that you should "not make a speech to impress the audience, but to tell a certain message, so you can impress the audience."

-Before you make a speech, plan, "I want to express something." My listeners want to know what to expect. ”

-Don't write down everything on your ppt, don't let your speech be full of tedious formula derivation that only interests you.

-When you ask questions, leave more time for them. Politely give the best answer when you answer a question. If you don't know the answer, say you don't know (say so). If you don't understand what the audience is saying, say so. Praise the audience when they say it is good or interesting. Write down the question and discuss the main purpose of your research is to get feedback from the audience.

-Pay attention to your voice and you should express your excitement and enthusiasm for the subjects you are studying. Be sure to focus on the important points. A weak voice will only tell the audience that you don't like your job.

-Make as many speeches as you can, and only practice will get better. 3.5 Write

-This is probably the most important part of the study.

-You have to be able to enjoy writing and be good at writing (if not, you will be in school for 4-5 years to achieve this goal). So you need to spend a lot of time here.

-This is the most original way you will tell the world about your research and communicate with the outside.

-Remember that even the best research, if not recorded, is equivalent to no user. A few years later no one knows it, remember it.

-In your research to fight for more productive. Remember that productivity is the production of quality and quantity.

-You have the ability to publish 4-5 papers per year as a single or co-author (2-3 short, 1-2 periodicals/long texts). In order to get an academic position that you look forward to, there are a number of papers in this way that can really quickly reach your goals.

-Write your paper in a place where you can concentrate, preferably alone.

-first list the outline of your paper, then add each section a little bit, and then add the details.

-Don't think about doing all the work before you start writing a paper, usually the process of writing helps you understand what you should be doing with your paper.

-Set deadlines to motivate yourself. Consciously write a paper before a certain date, or select a meeting to get a deadline for submission.

-When you're writing, think about the story you're going to tell. Make it interesting, not obvious, that the reader is justified by reading it. Check two times to make sure there are no obvious mistakes, let your mentor (or your co-authors) and other students read your manuscript and give you some constructive advice.

-Don't skimp on helping others. Even if you think someone has contributed very little to the paper, you should also ask him if he wants to be a co-author. The main author must be a large number of meaningful (thought, not insignificant) research parts or writers who have completed the paper. The author's name can be sorted alphabetically when everyone else is contributing to the same.

-Don't give up when your paper is rejected. This is always happening, even for the best researchers. Read the comments of the review, improve your work and post it again. Repeat this until received.

For a completely different job, you can write a technical report, deliver a short meeting paper or post a long paper to a journal. If you want your results to be repeated by the reader, you should package the relevant research content. Also, there is no need to publish the same work over and over again, and you should move on to your next job. 3.6 Programming

-This is the most basic ability you should achieve.

-If you don't have any other ideas, programming is enough to get you started on your research, you can come up with some interesting questions that deserve to be explored, build models for them, use programming simulations, and systematically study the effects of each parameter and get your own conclusions.

-You should have some basic programming language skills (Matlab,c,c++,java,perl scripting language), don't be afraid to choose a new language/development/development tool such as a network emulator. 3.7 Mathematical Analysis

-Programming is a way to study the network, but also the simplest way, undergraduates can do. You should learn to make your research rigorous by means of mathematical analysis that can be called graduate-level research.

-Choose a course with one or two real analysis (advanced or entry-level courses at the College of Mathematics), learn how to establish or prove theorems, and how to think Strictly.

-Learn to set up simple mathematical models for systems of interest to you, preferably starting with a simple first-order model and then spreading to more general situations. If necessary, add a layer of complexity and specificity to the model.

-Learn to identify interesting questions and use the math tools you have on hand to solve problems.

-Read the papers that contain complex mathematics carefully. For each theorem, make sure at least you understand what to prove and how to prove it, and what important assumptions are in the proof.

-Most importantly, don't be afraid when you're in a paper with complex math, what they need is just patience, understanding the patience of the meaning that the mathematical notation represents, understanding how those expressions are used, and why they are used so patiently. 3.8 background field knowledge

-Ask your mentor what course to choose.

-Don't go to class before you start the study. It's better to go to class once you know what your research needs.

-Select a thin field of study carefully. Understand this to overcome your weaknesses: for example, if you are very good at programming and computer systems, then take math. If you do not work on computer systems, then take a computer system. You can take a course to help you with your research. 4. The Art of research-the difficulty 4.1 Select a question

-The question should be very specific, and the problem you are studying should not be a broad concept.

-You may get some ideas from the paper you read. It's best to start by letting your mentor tell you a question. When you make perfect, you can choose a problem on your own.

-Once you've developed a sense of what you're interested in, ask the other person what's not, and try to figure it out in the reading.

-Always focus on problems rather than tools, and you'll find it more realistic and meaningful in the long run. 4.2 Formulation of a well-defined problem/model

-Start with some simple things, like some assumptions.

-Use tools you find useful, and use the tools at hand to formulate questions.

-as much as possible to express the problem in a specific formula. Try to express the problem more concretely, as in a postgraduate textbook (nontrivial).

-Learn how to create hypotheses so that your problems are neither very simple nor very difficult. 4.3 Thinking about the issues under study

-Make the problem a part of your life and leave the office as well.

-Ensure that the problem is accurately described and rested when it is jammed.

-In another way, consider a child problem and change the hypothesis.

-Try brainstorming methods to figure out how to solve the problem.

-Identify which tools are beneficial to the solution and make sure that you have accurate knowledge of the background. Don't be afraid to learn tools that might be useful.

-Try using analogical methods to connect questions to problems you encounter in a course or paper or discussion. 4.4 your mentor.

-You need someone to support you, mentor you during graduate school, and help you become a trained and mature researcher.

-There are two extreme types of mentors: a) hands-on. Give you a very good environment, that is, a project that can be studied, spend a lot of time with you, and supervise you until you finish the task. B) hands-off. Give you a lot of freedom, let yourself arrange, may just give a big research direction, occasionally meet with you. Both are good or bad. Type B: When you have a strong ability to drive, such mentors can give you the opportunity to identify research problems yourself. On the other hand, when you start a graduate student, you need to choose at least one person who can provide you with one or two experiences at the beginning of the study (the idea) to the end (the paper) to guide you. I personally think that the best mentor should do this: C. Flexibility to adjust your style based on your ability to drive, have some experience, and can be very loose (but will help you) and be able to control very tightly.

-See your mentor as often as possible, once or several times during the first week of your study, and increase or decrease the number of times you need to meet as needed. Don't meet the instructor for 2-3 weeks. Instead of meeting because you didn't make anything, it's best to meet regularly to motivate yourself to complete the task.

-Send an interesting paper or link to your mentor.

-Take the time to think about the questions you want to ask before the meeting.

-Choose different mentors at different stages of your academic career, and make progress as much as possible using this approach. (Help organize team meetings, do demos, make presentations, help write applications, and in the long run these can help you improve).

-In addition to your main mentor, try to find at least one or two other positions/fields of researcher to be your mentor, for example, a member of the Special Committee, or an intern unit. 4.5 Thesis

-There are different kinds of papers. Go to your departmental library to browse the latest papers in your field and have an idea of what's helpful to your thesis. (Note: In British English, the undergraduate and master thesis is called dissertation, and the doctoral thesis is thesis.) In American English, the undergraduate and master's thesis uses the thesis, the doctoral thesis uses dissertation).

-I advise you not to worry about your PhD thesis until you have worked on the project for several years (hopefully 2-3 years). Then choose a topic related to your research, a topic you are interested in, a topic that you have some early results and want to solve.

-even though a paper should be perfect to contribute to human knowledge, it is only when you realize that the doctoral thesis is not the end of your academic career but that you are ready to spend the rest of your life saving the world before you can begin to write your thesis.

Finally, while you'll certainly face stressful times,don ' t lose sight of the "big" picture. Graduate School is a great time for your to grow intellectually and personally. But this shouldn ' t come in the price of your emotional. Don ' t burn out completely. Make sure your give yourself plenty of time out to spend with friends and family. Keep everything in perspective. 5. Reference Documents

Books:
Stephen Covey, the Seven habits of highlyeffective people
Richard Reis, tomorrow ' s professor
Polya, how to Prove it
Antoine de St. Exupery, Little Prince
Jakob Nielsen, designing Web usability:the Practice of Simplicity

Internet Documents:
Ronald T. Azuma, so long and the PhD
Phil Agre, Advice for undergraduates considering Grad School
Phil Agre, networking on the network
Mary Desjardins, how to be a good graduate student
Wanda Pratt, graduate School survial Guide
Duane A.bailey, A letter to the Students
David Chapman, how do I at the MIT AI Lab