FAQ for Perspective Students   (Click here for Chinese  Version). If you feel like to join our lab or would like to know more about it after reading this FAQ, welcome emailing me to setup a meeting time. You can also go directly to LAB302 to chat with our students.

 

Q1:  Should I read this document?

A1: Yes, if you are interested in joining the "Machine Discovery and Social Network Mining" lab (MSLAB) as an undergraduate, MS, or Ph.D. student; or you simply have a lot of time to goof around.

Research: the most important thing you do as a graduate student

QR1: Is MSLAB a suitable place for you?

AR1: First of all, you have to be interested in our research. In a nutshell, we are doing intelligent information processing and I expect all the students in our lab to be very excited and enthusiastic about it. You can also propose your own research problem as long as it is along the line. Please check http://www.csie.ntu.edu.tw/~sdlin/mynews.htm for some existing projects and feel free to check the "popular science articles" about our research as below: http://www.csie.ntu.edu.tw/~sdlin/pscience.html

QR2: What kind of attitude do you expect?

AR2: you will need to have strong will to become an excellent problem solver and do world-level research. I will try my best to grow you into one of the best before you graduate, but it won't happen if you do not have the intension and perseverance to become one. We have seen many real world (e.g. Chien-Ming Wang) and fictional (One Piece) examples who succeed because of their will. You are encouraged to go through my talk slides to learn more about the right attitude of doing research. 
If you simply want to earn an easy degree and do not expect severe training, MSLAB is definitely NOT a good choice for you.

QR3: Should I have background in data mining, machine learning or information retrieval to be a student in MSLAB?

AR3: Not necessary as long as you are willing to learn. We will get you interested in them.

QR4: Do I need to work very hard to graduate?

AR4: Not necessary. You will be evaluated by your research results rather than how hard you work. It does not matter how many hours you spend in the lab. Work smart is more important than work hard. Nevertheless, we like responsible and earnest students and I feel it is my responsibility to teach you how to become such a person. 

QR5: Do I need to be good at programming and math to do well in MSLAB?

AR5: In general, people in our lab become good at programming or algorithm/math (or both).

QR6: Do I need to work on an NSC or industry project, or do I need to do system implementation?

AR6: Generally no, the only implementation you need to do is to implement the method you have designed to test how good it is.

QR7: What is the graduation criteria as a graduate student in MSLAB? How long does it take for me to graduate?

AR7:  For MS students, I expect a conference-paper level research quality. For Ph.D. students, you need to have journal-level contribution. Most of the MS students in our Lab have something ready to be published even before their final semester.
We will try to graduate an MS student in two years and Ph.D. student in 4 years, but no guarantee.

QR8: Do I need to find my own research problem or you will assign me one?

AR8: If you can find a favorite direction, great, I'll help you shape your idea into a doable one. If you cannot, then I'll provide some plausible topics for you to pick.

QR9: Is your work more theoretical or practical?

AR9: It is in between. We are bridging between theory and practice. We are not designing new algorithms and proving O(N) stuffs like you did in the algorithm course; neither do we purely implement systems like you did in the OS or compiler courses. We are doing the "modeling" work and we train you as a "problem solver". You will be given a real world problem to solve (e.g. how can you identify credit card fraud from a lot of credit card record? or how can you develop a system that can distinguish cancer X-ray images from normal images). To do such, you will need to design some method or algorithm and implement them to test the performance.

QR10: What kind of future road can an MSLAB student expect after graduation?

AR10: For MS students, the skills you learned in our lab will be welcomed in many software companies (e.g. Google, Yahoo, Cyberlink, IBM, TrendMicro, Microsoft, you name it). If performed well, you are also encouraged to keep doing research as a Ph.D. student (either in NTU or go abroad). For Ph.D. students, you can either join a school as a faculty member or start your industrial career in a variety of companies.

QR11: Do I do research on my own or team up with others?

Ar11: it depends on which way you like.

QR12: What is the official language of communication in MSLAB?

AR12: We speak in Mandarin/Taiwanese (and English for foreign students), but write in English.

Life is more than doing research

QL1: What can I learn in MSLAB besides how to do research?

AL1: Many things. MSLAB now has one faculty member, 4 Ph.D. students, 14 MS students, and 5-6 undergraduate students. You will learn paper writing, presentation, research ethics, Ultimate Frisbee, board games, bridge, playing online games, swimming, weightlifting, guitar, mountain climbing, taking care of little baby,  finding somebody who can potentially give birth to  your baby in the future, being a supportive friend, etc. We except your join will enrich the above list :)

QL2: Will I get funded?

AL2: Generally yes. The amount of funding will depend on how well you do on research. You will also be given some funding purchasing stationary and other things. You will be given a machine (desktop or laptop) to do research as well.

QL3: Do we have group or individual meetings?

AL3: Yes, we have group meeting once per week. You will meet with me individually every one or two weeks. There are also study groups you can attend.

QL4: Do MSLAB do sports, music or other entertainment?

AL4: Seeing is believing, please come to 302 to check it out.