Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Tips and notes by an early-stage researcher

Tips and notes by an early-stage researcher

These slides were used at the 2017 Advanced Research Methods course guest lecture, aimed at introducing artificial intelligence master's students to the field of academic research.

Suhas Vijayakumar

October 16, 2017
Tweet

More Decks by Suhas Vijayakumar

Other Decks in Research

Transcript

  1. Notes & Tips
    from an early-stage researcher
    Advanced Research Methods (lecture)
    Suhas Vijayakumar
    Oct 16, 2017
    http://suhasvijayakumar.in

    View Slide

  2. The promises

    View Slide

  3. The promises
    ● Behind the scenes - PhD
    Why do a PhD?
    How to get one?
    Oh! The chaos
    ● What after PhD?
    ● Why AI will change everything?
    ● Finding passion!

    View Slide

  4. Passion
    What’s that all about?
    Gillian Lynne

    View Slide

  5. Gillian Lynne

    View Slide

  6. Components
    Natural capacity.
    Terrence Tao
    You also love doing it.
    Ewa Laurance
    hated

    View Slide

  7. natural capacity
    is not a good enough
    reason to do it.
    Removed info

    View Slide

  8. Well, how do you know?
    Lose sense of time
    Gain energy rather than losing it
    You do it anyway, no matter the problems

    View Slide

  9. Well, how do you know?
    Birth Realization Goal
    Birth Realization
    mm..may be
    Goal

    View Slide

  10. Take risks

    View Slide

  11. https://www.youtube.com/watch?v=0KmimDq4cSU

    View Slide

  12. You take risks
    Bachelor’s
    Physics, Maths, Electronics

    View Slide

  13. 3 clues to understanding your brain
    https://www.youtube.com/watch?v=Rl2LwnaUA-k

    View Slide

  14. You take risks
    Bachelor’s
    Physics, Maths, Electronics
    Master’s
    Cognitive Neuroscience (research master’s)

    View Slide

  15. Daniel Wolpert’s TED talk
    The real reason for brains
    https://www.youtube.com/watch?v=7s0CpRfyYp8

    View Slide

  16. You take risks
    Bachelor’s
    Physics, Maths, Electronics
    Master’s
    Cognitive Neuroscience (research master’s)
    Ph.D

    View Slide

  17. What does it mean anyway?
    The illustrated guide to a Ph.D
    Matt Might | http://matt.might.net

    View Slide

  18. View Slide

  19. View Slide

  20. View Slide

  21. View Slide

  22. View Slide

  23. View Slide

  24. View Slide

  25. View Slide

  26. View Slide

  27. View Slide

  28. View Slide

  29. View Slide

  30. “Great… I want one

    View Slide

  31. So, you want to do a Ph.D
    Get your own money
    1st stream:
    funding from the government
    2nd stream:
    grants from NWO and KNAW
    3rd stream:
    other sources
    (universities, companies, charitable org...)

    View Slide

  32. Get your own money
    In general,
    1. Call for proposals
    (pre-proposal)
    2. Submission
    3. Review
    4. Defend your proposal
    5. Decision
    hellofalot of time | incredibly detailed | competitive
    https://www.nwo.nl/en/funding/our-funding-instruments/magw/
    research-talent/research-talent.html

    View Slide

  33. So, you want to do a Ph.D
    Get your own money
    1st stream:
    funding from the government
    2nd stream:
    grants from NWO and KNAW
    3rd stream:
    other sources
    (universities, companies, organisations)
    Look for a funded position
    option x:
    talk to friends/seniors/fb groups
    option x:
    look-up university websites
    option x:
    talk to your master’s thesis supervisor
    (or bachelor’s, or internship’s...)

    View Slide

  34. “I personally think that the most important thing
    during PhD is that you get a good chance to see
    what’s out there and you should find a position
    where you feel you have a good supervisor who’s
    willing to invest in you and where you can learn.”

    View Slide

  35. Organization of the motor cortex
    - Brain’s fundamental function = movement
    - Aspects of cognition
    - control of movement
    - perception of the environment
    - numerical cognition
    - Motor theory of language development

    View Slide

  36. Organization of the motor cortex
    What’s motor cortex all about?
    Identify on an anatomical basis the different circuits.

    View Slide

  37. Penfield and Boldrey (1937)

    View Slide

  38. Graziano (2016)

    View Slide

  39. Let’s study the frontal-parietal networks
    NOT SO FAST!
    First need to identify those regions
    based on data
    Averbeck et al (2009)

    View Slide

  40. Steps involved
    1. Region of interest
    2. Determine connectivity
    3. Cross-correlation matrix
    4. Employ a clustering algorithm (kmeans)
    5. Back-project to brain

    View Slide

  41. Start with a region of interest...
    The frontal lobe
    Hmph… what’s this region?

    View Slide

  42. Research happened!

    View Slide

  43. “Great… I want one

    View Slide

  44. It’s not always the “cutting-edge” of science!

    View Slide

  45. “Great… I still want one

    View Slide

  46. How to get a Ph.D?
    What are your interests?
    Demonstrate potential



    Meet them in person, if possible before applying
    Proofread your application and be yourself

    View Slide

  47. Be yourself

    View Slide

  48. Clip showing how the speaker conveyed teaching and research were
    equally interesting to him during the Ph.D interview.
    Point being -
    Ph.D is really like getting into a relationship!
    You don’t want to start on false beliefs and unrealistic hopes and assumptions.

    View Slide

  49. Aims
    this course
    ● Design, carry out and analyze
    empirical research in a
    methodologically correct way
    ● Critically read and evaluate
    academic articles
    ● Translate a theory into a practical
    framework, algorithm or a
    computational model
    ● Able to present the results in a
    scientifically rigorous way
    ● Collaborate with partners

    View Slide

  50. Year 0
    Start Ph.D
    Year 1
    Design 1st experiment
    Get some data
    Year 2
    Publish paper
    Go to conferences
    Design next n experiments
    Year 3 Year 4
    Get a Ph.D
    PUBLISH

    View Slide

  51. View Slide

  52. Start with a region of interest...
    The frontal lobe
    Hmph… what’s this region?

    View Slide

  53. Goulas et al (2016)
    While we were still working on things ...
    Our results (unpublished)

    View Slide

  54. View Slide

  55. Experiment 1
    Status: under progress… (until end of year 2)
    Motivation
    Status: under evaluation
    Future research
    Status: under construction
    Fix all the things!
    - Is NOT what you should do. Focus on your
    particular research question.

    View Slide

  56. Future
    1. Find what I love investigating
    2. See if it fits with the lab
    3. Design experiments
    4. Get a Ph.D
    5. Do all the things!

    View Slide

  57. General trajectory
    Ph.D
    Post-doc
    Post-doc
    Post-doc?
    Assistant prof
    Associate prof
    Full prof
    What next?

    View Slide

  58. General trajectory
    Ph.D
    Post-doc
    Mainly because
    - skill set/qualifications
    - need on-job training (sometimes)
    - inexperienced job in an unrelated field
    Switching can be difficult
    AI is an exception

    View Slide

  59. https://en.wikipedia.org/wiki/Applications_of_artificial_intelligence

    View Slide

  60. You really do have the opportunity to do
    Anything you want!

    View Slide

  61. The promises
    ● Behind the scenes - PhD
    Why do a PhD?
    How to get one?
    Oh! The chaos
    ● What after PhD?
    ● Why AI will change everything?
    ● Finding passion!

    View Slide

  62. Finding your passion
    Lose sense of time
    Gain energy rather than losing it
    You do it anyway, no matter the problems

    View Slide

  63. View Slide

  64. By no means easy
    Let’s give it a try anyway

    View Slide

  65. Questions? Exercise?

    View Slide

  66. Exercise
    1. Good at
    2. Enjoy doing
    3. Never tried
    4. Stopped doing
    5. Feeling

    View Slide

  67. I love doing
    I’m good at
    I can be paid for
    There is a need for
    PASSION MISSION
    JOB
    PROFESSION

    View Slide

  68. Questions submitted to Qn’A forum
    For more questions, feel free to contact
    @matteroftract
    (cue to remember: like “[as a] matter of fact”)
    Anonymous 11:24 AM
    "A PHD is more about becoming a better researcher than about the field"
    So, you think that the specialization doesn't matter too much or that after a PHD
    you can always come back to your "Passion" ?
    Upvote: 2 Downvote: 0
    Anonymous 11:51 AM
    How many hours a week do you spend?
    Upvote: 2 Downvote: 0
    Anonymous 10:54 AM
    How do you deal with the disappointments/feelings of insignificance that can arise
    during research?
    Upvote: 2 Downvote: 1

    View Slide

  69. goo.gl/forms/QxU0dh73coptYFx82

    View Slide

  70. Thank you
    Suhas Vijayakumar
    [email protected]
    http://suhasvijayakumar.in
    Oct 16, 2017

    View Slide