Presentation on theme: "Finding a Research Topic Padma Raghavan CSE Penn State With credits to: Mary Jane Irwin, CSE Penn State and Kathy Yelick, EECS UC Berkeley."— Presentation transcript:
Finding a Research Topic Padma Raghavan CSE Penn State With credits to: Mary Jane Irwin, CSE Penn State and Kathy Yelick, EECS UC Berkeley
The Thesis Equation Topic + Advisor = Dissertation
Area vs Topic Area = subfield architecture, theory, AI, high performance computing, or interdiscplinary Is it important? Timely? Jobs in the area? Topic = specific open problems in subfield Theory: provably better algorithm AI: Improving a machine learning algorithm Architecture: multicore cache design HPC: parallel algorithm, scheduling scheme Interdisciplinary: computer simulation of tumor growth
Topic Scale and Scope Scale Should have more than one open problem, or solving one should lead to another Should lead to more than one result/finding, some big, some smaller Scope Too narrow, e.g., just analysis no experiment, many not leave enough room Too broad, e.g., data mining, for what? why? too open ended
Passing exams Picking a Topic, Moving from coursework to research First publication Adapted from: Carla Ellis, Duke
Selecting a Topic Moving from coursework to picking a topic is often a low point Even for the most successful students Even for men (but they may not say so!) Why? Going from what you know-coursework, to something new-research! It is very important! There is no *one* ideal way, but many good ways
Selecting a Topic Is Important! It sets the course for the next two (or three) years of your life It will define the area for your job search You may be working in the same area (or a derivative) for years after It is uncommon to completely switch areas It is common to extend and add nearby areas
Things to Consider What kind of job are you interested in? Top-20 research univ, teaching, govt lab, or industry What are your strengths? Weaknesses? Programming, design, data analysis, proofs? Key insights vs. long/detailed system building, verification/simulation A combination? Narrow, broad, multidisciplinary ?
Topic vs Advisor Topic?= Advisor They are distinct but related choices At times hard to separate topic from advisor Interdisciplinary topic may need co-advisors, etc.
Things to Consider Do you have a preassigned research advisor or do you have to find one? How can your research be supported? By working as a TA By working as an RA for your advisor By having a university/college or NSF fellowship
More Things to Consider Does your advisor know anything about the topic? What is your advisors style? Are you more comfortable working as part of a team or alone?
1) A Flash of Brilliance You wake up one day with a new insight/idea New approach to solve an important open problem Warnings: This rarely happens if at all Even if it does, you may not be able to find an advisor who agrees
2) The Term Project + You take a project course that gives you a new perspective E.g., theory for systems and vice versa The project/paper combines your research project with the course project Warnings: This may be too incremental
3) Re-do & Re-invent You work on some projects Re-implement or re-do Identify an improvement, algorithm, proof You have now discovered a topic Warnings: You may be without a topic for a long time It may not be a topic worthy of a doctoral thesis It may be seen as incremental
4) The Apprentice Your advisor has a list of topics Suggests one (or more!) that you can work on Can save you a lot of time/anxiety Warnings: Dont work on something you find boring, badly-motivated,… Several students may be working on the same/related problem
5) 5 papers = Thesis You work on a number of small topics that turn into a series of conference papers E.g., you figure out how to apply a technique (e.g., branch and bound) to optimize performance tradeoffs Warnings: May be hard to tie into a thesis May not have enough impact
6) Idea From A B You read some papers from other subfields/fields Apply this insight to your (sub)field to your own E.g., graph partitioning to compiler optimizations Warnings: You can read a lot of papers and not find a connection Or realize someone has done it already!
* … Combine, compose Try any combination of these ideas But, focus on tangible progress, milestones Warnings: It can take a lot of time without any results!
Some Tips Research topic and advisor are both important Keep an ideas notebook; these could turn into research papers later Follow your interests and passion Key driver for success and impact Are you eager to get to work, continue working? If not really interested, correct and adapt But, differentiate between tedium versus real lack of interest and motivation
Set Goals/Take Stock Set goals for a topic-finding-semester E.g. Selecting and trying 2 of 6 strategies Assess your progress Are you converging to an area? Or have you ruled out an area? Have you got a workshop paper or term project+ done? Adapt your strategy
When Youre Stuck …. Serve as an apprentice to a senior PhD student in your group Keep working on something Get feedback and ideas from others Attend a good conference on a hot topic http://www.cra.org: Grand challenge conferences, CRA-W Summer Schools http://www.cra.org Do a industry/government lab internship
When Youre Stuck … Read papers in your area of interest Write an annotated bibliography Present possible extensions/improvements to each Read a PhD thesis or two (or three) Attend oral exams, thesis defense of others students Read your advisors grant proposal(s)
Take Risks ! Switching areas/advisors can be risky May move you outside your advisors area of expertise You dont know the related work You are starting from scratch But it can be very refreshing! Recognize when your project isnt working It is hard to publish negative results
Take Risks ! Take some risks in your research Choose problems that are significant Higher risk to solution Higher reward for solution But, balance High risk ---may not have solution, negative results cannot be published
Find a Topic and Forge Ahead! Questions Comments Discussions