Wenfei Xu

@iamwfx

Urban planning PhD student | Housing, redlining, "data stuff" | Previously: and Spatial Data Scientist | She/her

New York, NY
Vrijeme pridruživanja: studeni 2012.

Tweetovi

Blokirali ste korisnika/cu @iamwfx

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @iamwfx

  1. Prikvačeni tweet
    30. kol 2018.

    Finally got around to finishing this map!: Looks at the relationship between zones and socioeconomic indicators across 147 US cities between 1930 and 2016. Compares 's georeferenced maps with a spatial join of census data.

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    6. velj

    Gus Wezerek and I have an editorial in today's New York Times discussing potential impacts of the census bureau's new privacy algorithm on 2020 decennial census data.

    Prikaži ovu nit
    Poništi
  3. proslijedio/la je Tweet
    3. velj

    OMG: phone wagon traffic jam guy wasn't trying to 'hack Google Maps,' he was trying to show that "the map is not the territory" and “thus data are viewed as the world itself, forgetting that the numbers are only representing a model of the world” :swoon:

    Prikaži ovu nit
    Poništi
  4. 4. velj
    Poništi
  5. proslijedio/la je Tweet
    29. sij

    Mental note to read new report by a (long) list of very smart people conceptualising Urban Science and pointing ways forward. Looks thought provoking and great to estimulate discussion:

    Poništi
  6. 23. sij

    It's that time of the year again! does an annual street homeless population count and it's happening Mon (Jan 27) evening at 10pm in all five boros. If you have a late evening/early morning to spare, sign up to volunteer👇

    Poništi
  7. 17. sij

    Of course, ask me in a few weeks about this. I'm sure it will be going, uh...swimmingly. 😅✌️

    Prikaži ovu nit
    Poništi
  8. 17. sij

    I don't hide the fact that coding can be frustrating. There's a lot of hand-wringing and Googling involved (a key lesson). However, I emphasize that the fundamentals of coding or ML are not _conceptually_ difficult. It's just your computer running stuff over and over again. 11/

    Prikaži ovu nit
    Poništi
  9. 17. sij

    And because these students are mostly social innovation designers (visual learners, like me), we’ll be talking a lot about “the diagram” and probably little or no math. 10/

    Prikaži ovu nit
    Poništi
  10. 17. sij

    Also, no github, I'm sorry to say. Do most people really use it unless they're professional developers or need to collaborate? I find that it typically causes more confusion at the beginning of a course (when it's typically taught). Google drive all the way! 9/

    Prikaži ovu nit
    Poništi
  11. 17. sij

    Thankfully, there are so many wonderful python/pandas resources out there. (I don’t quite get why people write their own?) We’re using a combination of Python Data Science Handbook, , and to a lesser extent, Python for Data Analysis. 8/

    Prikaži ovu nit
    Poništi
  12. 17. sij

    So, we're using Google colab. No set-up and in an environment students are already familiar with. 7/

    Prikaži ovu nit
    Poništi
  13. 17. sij

    First, this xkcd comes to mind. When I was first learning how to code, I spent a solid few weeks just figuring out how to set up all the dependencies I needed (not even talking about geospatial libraries here…) 6/

    Prikaži ovu nit
    Poništi
  14. 17. sij

    How can I cultivate these talents in my students and lower the barriers to entry to a field that is way over-hyped? 5/

    Prikaži ovu nit
    Poništi
  15. 17. sij

    Case in point: I showed a diagram of k-means clustering in class last week. One of my students asked me if it’s just a matter of minimizing distances between points. So smart! 4/

    Prikaži ovu nit
    Poništi
  16. 17. sij

    It wasn’t until grad school (the first time), where I took my 1st ML course, that I realized I’m probably just a visual learner. Each algorithm/proof was first presented as a picture before we went into the math. And (not so) shockingly, these concepts aren’t that difficult! 3/

    Prikaži ovu nit
    Poništi
  17. 17. sij

    I took a lot of theoretical math and econ in college, but I was also intimidated a LOT. I really enjoyed the comp sci class I _started_ but felt too incompetent to finish. I also thought everything had to be expressed with⚡️rigor⚡️, meaning symbols and proofs. Sigh, college. 2/

    Prikaži ovu nit
    Poništi
  18. 17. sij

    I’m just starting to teach a new “data science” course (it’s just python/pandas/ML) at , so I’ve been thinking a lot recently about how to teach technical tools to a mostly non-technical audience and also! how to create a supportive learning atmosphere. (thread) 1/

    Prikaži ovu nit
    Poništi
  19. proslijedio/la je Tweet
    17. sij

    computer but make it fashion

    Poništi
  20. proslijedio/la je Tweet
    13. sij

    Just one week left to submit your project proposal for Data Through Design 2020! Our theme this year is "Digital Twin". Find out more and submit your project idea here:

    Poništi
  21. 16. sij

    We'll be hearing about smart enough cities, digital twins, gentrification and displacement, NIMBY/YIMBY, data feminism, sensing inequalities, ML and urban boundaries, +++. Come through if you're in NYC!🥰

    , , i još njih 2
    Prikaži ovu nit
    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·