CS Independent Work Tips for effective communication Getting - PDF document

Outline • What is independent work/research? • Schedule and course work CS Independent Work • Tips for effective communication Getting Started • Summary • Faculty presentations Randy Wang Fall, 2002 Princeton University Independent Projects 1 Randy Wang Outline What Is Independent Work? • What is independent work/research? • Exposure to research • Schedule and course work • Opportunity to work with faculty and grad students • Tips for effective communication • So, what is research? • Summary - Formally: advance state of art • Faculty presentations - Informally: tell people something new Independent Projects 2 Randy Wang Independent Projects 3 Randy Wang

What Is Research and What Is Not? What Is Research and What Is Not? (cont.) • Non-research - My advisor gave me this mpeg decoding algorithm • Research - I implemented it - My advisor gave me this mpeg decoding algorithm - And it worked - I implemented it - I measured it • Research - I analyzed its bottleneck - I took two existing mpeg decoders - I instrumented the code to prove the hypothesis - I took some sample movies - I recommend and conclude - I studied the decoders qualitatively - I measured them quantitatively • Research - I concluded why one is better - I was given an mpeg decoding implementation - I identified its bottleneck as above - I proposed an improvement - I implemented the improvement - I measured it again to prove/disprove I’m right - I generalize and conclude Independent Projects 4 Randy Wang Independent Projects 5 Randy Wang What Is Research and What Is Not? What Is Research? (cont.) • Research Research = - My advisor asked me to implement an mpeg decoder Analysis + Synthesis + Hypothesis - I came up with 2 alternative designs - I analyzed qualitatively why one might be better - (I experimented to directly or indirectly deduce I was right) Him/her Them Me - I implemented the best design - I summarize and conclude • Research - Many other possibilities... Literature Innovate • So, what is research? Survey - Formally: advance state of art - Informally: tell people something new - Not necessarily much much more work - Just need to “go the extra mile” Independent Projects 6 Randy Wang Independent Projects 7 Randy Wang

Other Traits of a Good Project Outline • What is independent work/research? • Interesting/important problem • Schedule and course work • Non-trivial challenge(s) • Tips for effective communication • Exploration of new technology • Summary • Can be finished in allotted time • Faculty presentations • Effective communication (talks, reports) Independent Projects 8 Randy Wang Independent Projects 9 Randy Wang Who Does Independent Work? Independent Work Schedule (cs397, cs497, csJIW) • AB students • By 10/1: Project plan - Junior independent work (JIW) for two semesters - Find an advisor and a project - Senior independent work (SRT) for two semesters - Fill out an “Independent Work Project Form” - Automatically registered - Register on-line • BSE students • 10/14 and 10/16: Project proposal talks - At least one semester during junior or senior year - 5 minute presentation to students and prof. - Encourage more than one semester • During the project (2nd semester counts a departmental) - Work 10-15 hours per week - Juniors: cs397 in fall, cs398 in spring - Meet regularly with your advisor - Seniors: cs497 in fall, cs498 in spring • 11/18: Project checkpoint - 4-5 slides report progress and remaining plan • 12/9-12/12: Project results talk - 13 minute presentation to students and profs. • 1/6: Final written report Independent Projects 10 Randy Wang Independent Projects 11 Randy Wang

Independent Work Schedule Find an Advisor and a Project (csSRT and BSE Thesis) • Get info about profs’ research • By 10/11: Project plan - Independent work page, home pages, research - Find an advisor and a project papers, word of mouth, ... - Fill out an “Independent Work Project Form” • Schedule meetings with several profs - Register on-line - email, office hours, appointments • 10/23: Project proposal talks • Choose a prof - 8 minute presentation to students and prof. - Must be from CS dept. • During the project - Can be jointly advised by someone outside - Work 10-15 hours per week • Decide on a project - Meet regularly with your advisor - Profs suggest choices • 11/20: Project checkpoing - Students come up with their own - 4-5 slides report progress and remaining plan - A combination • 12/9-12: Project progress talk - Mutual agreement, interest, enthusiasm - 13 minute presentation to students and profs. • Submit “Independent Work Project Form” and • 1/6: Written progress report register on-line • May: Submit thesis Independent Projects 12 Randy Wang Independent Projects 13 Randy Wang Project Proposal Talk Project Proposal Talk and Beyond • Scope • Problem description - Not too little - What am I going to do - Not too much (carve out a subpiece, limit - Why is it important functionality, reduce measurements) - If you’re ambitious, have a longer term plan but the - Why is it hard short term plan should still be doable • Approach - Don’t be afraid of negative results - Previous approaches - Don’t be vague - My approach • Have a theme - Why is mine better - What is the core/new/clever/hard/cool nugget? • Methodology, milestones, deliverables - Repeat it - Specific steps - Make sure everything else serves the theme - What steps/deliverables will be done by checkpoint • Be conscientious - What other steps/deliverables will be done by end of - Start early semester - Define small milestones for yourself - What might be hard and what’s the fall-back plan - Work continously to meet milestones • Summary - Don’t hesitate to get help Independent Projects 14 Randy Wang Independent Projects 15 Randy Wang

Project Checkpoint End-of-Semester Talk • 4-5 slides • Review the problem description and proposed approach---give “the theme” • What you proposed to have done by checkpoint • Give details (eg., of implementation) that are • What you have actually accomplished by relevant to “the theme” checkpoint • Give key results to support “the theme” - Steps - Deliverables • Summarize “the theme” • Difficulties/surprises/deviations? • What more do you do you expect to do - Steps - Deliverables Independent Projects 16 Randy Wang Independent Projects 17 Randy Wang End-of-Semester Report Grading • Introduction • (Proposal talk: 8%) - Background • Checkpoint: 7% - Problem description: include goal • End-of-semester talk: 15% • Approach • End-of-semester report (may include - Previous approach(es) “participation”, “draft”, etc.): 70% - My approach - Why is mine better • Detailed description of methodology or • No grade inflation! implementation • All steps count • Experimental results - Analyze/interpret data, don’t just give numbers - What does this have to do with your theme? • Conclusion • Acknowledgements and bibliography Independent Projects 18 Randy Wang Independent Projects 19 Randy Wang

Outline Effective Writing • What is independent work/research? • Concepts - Focus on key ideas • Schedule and course work - Expose, weave it in the body, conclude • Tips for effective communication - Why, not what • Summary - Don’t get trapped by details and artifacts • Faculty presentations • Flow: be a good story teller - Pay attention to order, and smooth transitions • Simplicity - Simple exposition, simple styles - Make it intuive rather than formal - Be specific, use examples • Pitfalls - Too vague: for example, no examples - Contributions none-obvious - Too much “what”, not enough “why” Independent Projects 20 Randy Wang Independent Projects 21 Randy Wang Ten Commandments of Talk Tips - Edit Giving a Bad Talk • Focus on key ideas • Choose carefully a sub-story out of a full story • Thou shalt not be neat • Understand the nature of oral communication (no pause, no rewind) • Thou shalt not waste space • KISS (keep it simple, stupid) • Thou shalt not covet brevity • Repetition is useful • Thou shalt cover thy naked slides • Give outline: some amount of predictability is comforting • Thou shalt not write large • Explain, interpret, justify, not just describe • Thou shalt not use color • Write large • Thou shalt not illustrate • Use color, but sparingly, consistently • Thou shalt not make eye contact • Use pictures (and even animations) • Thou shalt not skip slides in a long talk • No full sentences (just terse outline) • Make the sub-story coherent and self-contained • Thou shalt not practice • One corollary: no standalone graphs Independent Projects 22 Randy Wang Independent Projects 23 Randy Wang

Talk Tips - Preparation Talk Tips - Presentation • Practice • Speak clearly, make eye contact • Dry run followed by slide-by-slide analysis • Don’t read slides • Pay attention to time and practice for time • Pay attention to posture • Practice for varying audience backgrounds • Eye contact, shift gaze • Plan on shedding slides • Admit shortcomings, don’t wait for questions • Analogies and jokes can help Independent Projects 24 Randy Wang Independent Projects 25 Randy Wang Outline Summary • What is research? • What is independent work/research? - Teaches people something new • Schedule and course work • Course work • Tips for effective communication - Be conscientious and stick to a plan • Summary • Effective communication - Stick to a theme and tell a good story • Faculty presentations Independent Projects 26 Randy Wang Independent Projects 27 Randy Wang