Julia Silge

@juliasilge

Data science and modeling at , , text analysis, parenthood ✨📊✨ she/her

Salt Lake City, UT
juliasilge.com
Geregistreerd in februari 2008
2.160 foto's en video's Foto's en video's

Tweets

Je hebt @juliasilge geblokkeerd

Weet je zeker dat je deze Tweets wilt bekijken? @juliasilge wordt niet gedeblokkeerd door Tweets te bekijken.

  1. Vastgemaakte Tweet
    3 feb.

    I am very happy to announce that 💫finally💫 I have published a free interactive course for learning tidytext! This course uses learnr and walks through four case studies, from TED talks 🗣 to Shakespeare 🎭 to the NYT 🗞 to song lyrics 🎵

    29 antwoorden 533 retweets 2.220 vind-ik-leuks
    Ongedaan maken
  2. 22 jul.

    More ! Hoping to make plots, train models, and have some fun 💞

    The #SLICED logo and the following text with headshots of each contestant on a swirling pink, turquoise, and purple background FOUR CONTESTANTS advancing to the #SLICED playoffs based on votes from chat and the top four contestants! David Robinson, Ethan Douglas, Julia Silge, Josh Pohlkamp-Hartt twitch.tv/nickwan_datasci
    And congrats to our next four contestants advancing to the playoffs based on votes from chat and the top four contestants!!! !
    Deze collectie tonen
    1 retweet 35 vind-ik-leuks
    Ongedaan maken
  3. heeft geretweet
    21 jul.

    I cannot overemphasize how true this statement is. I spent the 4 weeks before started seriously learning ML on nights and weekends using {tidymodels}. not only was I ready to compete, I made it to the playoffs. the packages are THAT GOOD y'all.

    1. The tidymodels package is a really powerful "pit of success" for machine learning 🔥 Perhaps the most important innovation is the combination of recipes (feature engineering) and models into an encapsulated workflow. (Continued)
    Deze collectie tonen
    1 antwoord 8 retweets 107 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  4. heeft geretweet
    21 jul.

    NEW BLOG POST: Machine learning in a hurry: what I've learned from the ML competition

    Meme with text: Planes labeled with "LINEAR MODELS WITH TRANSFORMATION, REGULARIZATION, DIMENSIONALITY REDUCTION, AND HAND-CRAFTED FEATURES AND INTERACTION TERMS" Superheroes labeled with GRADIENT BOOSTED TREES: Look what they need to do to mimic a fraction of our power
    9 antwoorden 84 retweets 477 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  5. heeft geretweet
    20 jul.

    Tonight is the night! This is the final episode of 🔪📊 Lap 2 before we head into the playoffs! Watch at 8:30PM ET on Who are you going to vote for??? Will it be wilder than last week???

    SLICED logo with the contestants names and twitter handles on a bright green background with colorful illustrations that look like skyscrapers Landon @magi_param Scott @scottrcole Julia @juliasilge Tony @tonyelhabr
    1 antwoord 4 retweets 12 vind-ik-leuks
    Ongedaan maken
  6. heeft geretweet
    20 jul.

    readr 2.0.0 is now on CRAN 🎉 This is a _major_ release of readr that adds built-in support for reading multiple files at once, fast multi-threaded lazy reading and automatic guessing of delimiters among other changes. See for details on this release!

    15 antwoorden 136 retweets 617 vind-ik-leuks
    Ongedaan maken
  7. heeft geretweet
    20 jul.

    I've seen a lot today about how effective the vaccines are; mistakes aside, lots of folks seem to be mixing up which denominators matter - good thing statisticians *LOVE* denominators 🥰 1/6

    37 antwoorden 1.258 retweets 2.554 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  8. 19 jul.

    This time tomorrow night, I'll be getting ready to do this again!!! 😬

    Julia Silge I have three kids, and my favorite author is Jane Austen. I spend my leisure time on reading, exercise, and excessive skincare routines. I just got a new Mac Mini and I’m figuring out how to do data science on the new M1 chip. No surprises, I’m using the RStudio toolset for SLICED. I transitioned to data science in my mid-30s!
    Landon Buechner Outside of all things academic I like to run long distance, read epic fantasy/scifi, drink copious amounts of coffee, and drink IPAs with the boyz. My rig’s specs are: i7 9700k, 1tb Samsung EVO SSD, 32 GB RAM, Nvidia 2070 GPU. But I used Colab in the pilot season of SLICED. I’m a recent graduate and ready to use my skills in the real world.
    Scott Cole Every morning, I wake up at 6am to wander the city while it's still empty, listen to podcasts, and feel the effects of caffeine. My setup is pretty basic. For SLICED I’m using a Jupyter notebook on a MacBook Pro my PhD advisor bought me. Though I have a "real job" now, I still like to live like a grad student. Each day I eat 1-2 clif bars and a can of chickpeas.
    Tony ElHabr Outside of data science, I like playing basketball, reading, meme-ing, piano, listening to music, and boardgames. I’ll be using #rstats, tidyverse, tidymodels. All on an 8GB RAM laptop to torture myself. I’m best at data visualization and storytelling, okay at data engineering, and least skilled in machine learning. Since the SLICED pilot, I still haven’t learned to ride a bike.
    The final episode of 🔪📊 lap 2 is TOMORROW starting @ 8:30PM ET at ! Featuring: Then next week we kick-off the first of two playoff rounds before the final four and championship episodes! 🏆🏆🏆
    5 retweets 27 vind-ik-leuks
    Ongedaan maken
  9. 19 jul.

    We are *so* lucky to have Jesse joining the open source team! 💕

    y’all, today I’ve joined the open source team as a Developer Advocate, alongside the inimitable 👯 one of my first areas of focus will be getting up to speed with Shiny, and I can’t wait for us to learn and grow together ✨🌱💕
    2 antwoorden 1 retweet 80 vind-ik-leuks
    Ongedaan maken
  10. heeft geretweet
    17 jul.

    How should we feel about that deepfake Anthony Bourdain voice in “Roadrunner”? I talked to Morgan Neville, Ottavia Busia, and some seriously smart A.I. ethicists about what happened, what didn’t happen — and what might happen

    26 antwoorden 129 retweets 552 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  11. 17 jul.

    At least I made and ate dinner somewhere in there

    6 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  12. 17 jul.

    If you decide to write a book, YOU TOO can spend your evening doing this kind of really interesting and delightful work

    GitHub screenshot of commit messages, just a whole bunch of "copyediting" and "more copyediting" and "yet more copyediting" of REFERENCES, not even anything interesting
    8 antwoorden 4 retweets 120 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  13. 16 jul.

    🎯I am back on this coming Tuesday evening for the last episode of the 2nd round!!! Super happy about the points changes since the 1st round and feeling excited. 🗳Do I want you to vote for me to get to the finals??? Honestly I do not know🙈but I appreciate the love. 💕

    CURRENT STANDINGS VOTING ENDS WEDNESDAY WHEN I GO LIVE (8:30PM ET) SEND YOUR FAVS TO THE PLAYOFFS. TOP 2 GO THROUGH (IF THEY AREN'T ALREADY IN)
    2 retweets 28 vind-ik-leuks
    Ongedaan maken
  14. heeft geretweet
    16 jul.

    There ARE ethical ways to create a synthetic voice from a donor voice, but it requires clear consent from the donor. The voicebank is a good example.

    4 retweets 21 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  15. 16 jul.

    Um, the last thing I just RT'd! 🤯 If you didn't read the New Yorker piece by yet and you work with people's data and machine learning, worth a look.

    Kinda wish I hadn’t written the Bourdain AI thing so that I could write a piece weighing in on the Bourdain AI thing
    Deze collectie tonen
    10 vind-ik-leuks
    Ongedaan maken
  16. heeft geretweet
    16 jul.

    I certainly was NOT the one who said Tony would have been cool with that.

    "I checked...with his widow and his literary executor, just to make sure people were cool with that," Neville said. "And they were like, Tony would have been cool with that. I wasn’t putting words into his mouth. I was just trying to make them come alive.”
    Deze collectie tonen
    220 antwoorden 4.311 retweets 29.186 vind-ik-leuks
    Ongedaan maken
  17. heeft geretweet
    15 jul.
    Als antwoord op

    Thanks! But I’d like to give the credit to the students in the capstone course taught by ! Check out their letter to the Journal of Statistics & Data Science Education!

    4 retweets 10 vind-ik-leuks
    Ongedaan maken
  18. 15 jul.

    Thanks so much to for his thoughtful discussion on this! 💕

    There's a new version of the {{modeldata}} package on CRAN. There is a great data set (tate_text). We're also going to remove the two OkCupid data sets in the next version.
    1 antwoord 1 retweet 15 vind-ik-leuks
    Ongedaan maken
  19. 15 jul.

    HERLOCK. SHOLMES.

    4 antwoorden 15 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  20. 15 jul.

    I'm finishing up the second season of Lupin, and I have now learned about Herlock Sholmes, one of the best things to come to my attention in A WHILE:

    4 antwoorden 3 retweets 35 vind-ik-leuks
    Deze collectie tonen
    Deze collectie tonen
    Ongedaan maken
  21. heeft geretweet
    14 jul.

    In tonight's , gonna be rocking the development versions of parsnip & workflows A. parsnip models now have a default instead of needing set_engine() B. workflow() now takes a preprocessor / spec argument 💕 the recent work from the tidymodels team

    # Screenshot of the following code devtools::install_github("tidymodels/parsnip") devtools::install_github("tidymodels/workflows") library(tidymodels) lin_wf <- recipe(mpg ~ wt, data = mtcars) %>% step_log(wt) %>% workflow(linear_reg())
    2 antwoorden 12 retweets 78 vind-ik-leuks
    Ongedaan maken

@juliasilge heeft nog niet getweet.

Het laden lijkt wat langer te duren.

Twitter is mogelijk overbelast of ondervindt een tijdelijke onderbreking. Probeer het opnieuw of bekijk de Twitter-status voor meer informatie.

    Je bent misschien ook geïnteresseerd in

    ·