THE 6TH EDITION OF THE DSSG BERLIN DATATHON IS COMING - SIGN UP NOW🤘The 6th edition of our #datathon is taking place #remotely on March 24-26. Sign up NOW as participant, organizer #volunteer, or #data ambassador! See you soon!🤩 https://t.co/5Hmyjuy6l7 #DataScience#Hackathon
Fantastic talk by @HannahBohle at @PyConDE about transitioning from science to data science. The tips she shared are really valuable and provide a much needed guidance for making this career leap. #PyladiesAtPyConDE2022
Thanks for the nice feedback! Great audience today :-) In case anyone else wants to catch up on relational databases and the query planner, slides are here: https://t.co/m2Kj47CL52
Somewhat comfortable with using SQL to access data, but curious to know what happens behind the scenes when you send off your query? Join my talk at #PyDataBerlin tomorrow! https://t.co/8JCNELmeQZ #PyConDE
@a_galgal @tamarajul Yes, if one happens to visit the UK, I can only recommend doing the personal fitting. Very professional, warm experience, not awkward at all.
Hyreskartan visar hur många hyresrätter som förmedlades genom Stockholms Bostadsförmedling under 2020. Använd kartan för att kolla om/var du har chans att få lägenhet. #bostadsförmedligen https://t.co/XBuypM3wYY
Asynchronous execution: running inference can take a long time and you don't want to lock up the browser tab while waiting for the results -> tensorflow.js heavily uses async execution. Again, tutorial has examples, which I hope explain the concept to people unfamiliar with it.
I recently played around with tensorflow.js -- used it to generate (in the browser!) adversarial images (=images that fool computer vision models into making wrong classifications). Tutorial with live demo: https://t.co/ZLFaHN7e3K
Memory management: you need to manually "free" any tensors that tensorflow.js allocates for you. Otherwise: memory leaks, ouch. Bit of a hassle + tf.tidy has a daunting syntax... hopefully my examples help understand what is going on.
@xoryouyou @hnshck Hi Till. Ich kenn mich auch ganz gut aus im Bereich Computer Vision. Hans und ich wollen die Tage mal telefonieren, ob/wie man sowas hier einsetzen kann. Magst du dabei sein?
@zweifeln@xHain_hackspace@cbase@afra_berlin Bisschen weiter draußen (in Buch): https://t.co/iTRCScNhDW nicht kostenlos, aber dafür sehr nette persönliche Betreuung. Ich fands da super.
My blogpost is done! 😄 How does our perception shape how we work together across different backgrounds, and expertise? How can we work together better?
https://t.co/MgbpzvQjJr
@ellen_koenig @terezaif @marielli @kycl4rk @pythonpizzaconf I have some unorganized thoughts on testing that data conforms to expected specs. Maybe that is not really a unit test, though.
@ellen_koenig @terezaif @marielli @kycl4rk @pythonpizzaconf I gave a presentation about this topic some years ago - it ended up being more about unit testing in general, less on the data science. It is also pretty outdated, I would use pytest now for the examples instead of nose. But maybe it helps anyway: https://t.co/O2h3UfGkcm
@floris_du_mal @PoLILYtik Das ist nicht korrekt. Diese Richtlinien wurden beschlossen für den Fall, dass die Intensivbetten nicht mehr ausreichen (Fall noch nicht eingetreten). Allerdings gibt es Hinweise, dass Richtlinien trotzdem teilweise befolgt wurden, wird aktuell untersucht: https://t.co/ENprJwnzsz