Alexander Schäfer (@AlexandSchaefer) profile picture

Former neuro scientist @maxplanckpress and @NUSingapore. Now data scientist @Cape_Analytics

Berlin, Deutschland
Joined March 2013
Photo by Alexander Schäfer (@AlexandSchaefer)
Students are not used to iterative work processes. Please keep in mind when you work with them on their thesis.
After advising PhD & Master students for over a decade, there is one thing I find most students need to unlearn: the half-ass work mentality acquired during years of tests and homework. Let me explain (thread 🧵). 1/N #AcademicTwitter
Free access to ACM Digital Library during coronavirus pandemic: acm.org/articles/bulletins/2…
Its 2019 and #Google gives you a Tesla K80 for free lnkd.in/dhMVNkd. So the pain of installing or updating #CUDA drivers on my laptop are a thing of the past. lnkd.in/dEwjw_X
1
1
Are we in a data science supply bubble? Great read by Vicki Boykis with a lot of good recommendations for junior people and students. lnkd.in/drv4Zyd lnkd.in/dSqwVzZ
2
Let's Make #Artificial #Intelligence 'Boring' Again. #AI is just one tool it's just as exciting as your C compiler. https://teapaw.org/t.co/2EVeP2dOEH
1
If you often solve similar problems with #ML you can safe a lot of computing time and data using #meta-learning. The great people from @fastdotai published an algorithm and code solving this very elegant. lnkd.in/dxBQYGA
2
What If... you could inspect a #ML model, with no coding required? Amazing extension of #tensorboard especially for teaching and presenting models. Question might be if it scales? goo.gl/wiUKWR
1
3
1
13
If you are working with #jupyter #notebooks you should have a look at Reproducible Data Analysis in Jupyter by Jake Vanderplas. Covers many important topics like visualization, #packaging and unit #testing with jupyter goo.gl/UxL79t
3
5
Great slides by Andrew Zisserman on the rising topic of Self-Supervised Learning in #ML project.inria.fr/paiss/files…
2
6
Next time #turing #completeness is sold you as strong feature. Refer to software that is turing complete by accident beza1e1.tuxen.de/articles/ac…
2
Deep clustering becomes relevant on high dimensional (visual) data arxiv.org/abs/1807.05520
1
I can recommend listening to this. Andrew Ng gives a nice high level/ business perspective on AI youtube.com/watch?v=NKpuX_yz…
1
3
arxiv.org/abs/1802.06430 adversarial attacks on #autonomous cars via manipulated traffic signs
1
Ink to code. Microsoft already has an app that translates sketches into code. microsoft.com/en-us/garage/p…
1
1
Why ML Companies Can’t Be Lean Startups https://teapaw.org/t.co/dKarUutTpb