Hannah Bast

[Hannah Bast @hannahbast]

@srchvrs One advantage of Naive Bayes: it's a simple and fast classification method (training is essentially just counting vs. an iterative method like for the otherwise superior multinomial regression)

[Hannah Bast @hannahbast]

Average height by occupation and gender, based on Wikidata

https://t.co/Amshhvlpju

For example, female volleyball players are taller than male golfers (on the average).

#aqueryaday

[Hannah Bast @hannahbast]

@lintool 2/ If you write larger amounts of code without extensive unit tests, it will not be correct. If you publish complex results without a precise replicability setup, the results will not be correct.

[Hannah Bast @hannahbast]

@lintool 1/ Drummond's main argument is that replicability "would serve as little more than a policing tool, preventing outright fraud." He misses the point. The problem is not fraud, the problem is correctness.

[Hannah Bast @hannahbast]

@srchvrs My experience with ChatGPT is that in an actual conversation, you don't mind waiting for a few seconds if the answer is good. One Q&A cycle is then more like a sequence of multiple Google searches

[Hannah Bast @hannahbast]

Elizabeth Holmes' defense ("she was very young and deluded") reminds me of this classic

[Sebastian Ruder @seb_ruder]

10 Things You Need to Know About BERT and the Transformer Architecture That Are Reshaping the AI Landscape

This super comprehensive post covers most things that are important in current NLP including BERT, transfer and avocado chairs 🥑
by @cathalhoran

https://t.co/Hn458pkrk8

[Sharif Shameem @sharifshameem]

Found this comment in Apple’s open source chess engine source

[Hannah Bast @hannahbast]

I love it ❤

[Leo Boytsov @srchvrs]

@annargrs 👇It was very interesting to read your carefully written opinion piece. Coincidentally, there is another non-trivial NIPS-style experiment and it is published. It doesn't go too much in the description of biases, but main conclusions align well:
https://t.co/UbvpN6QTy3

[Hannah Bast @hannahbast]

The ESA 2018 Track B Experiment: An in-depth analysis of two parallel program committees reviewing the complete set of submissions independently of each other. Read the slides, the results are quite interesting
https://t.co/jWGhMwJPgg

[Mario Brandenburg @BraFDP]

Ein Traum wird wahr!
Endlich logische Diskussionen im Deutschen Bundestag 🤣

#EnqueteKI #bundestag #logik #informatik #ki #ai #künstlicheintelligenz #artificialintelligence #fdp #brafdp

[Hannah Bast @hannahbast]

@arjenpdevries Wow :o and I recognize the hall, I gave a keynote in it two months ago. It wasn't nearly as full though :-|

[RiseML @riseml]

Google Cloud TPU as fast as 4x Nvidia V100 while being 57% cheaper on ResNet-50. Read our new benchmark article. https://t.co/bAMoSVfIVx

[RiseML @riseml]

📊 First benchmark on Google's new TPUs for training @RiseML https://t.co/MYGfeWrxzq #google #nvidia #tpu

[RiseML @riseml]

Reduce your cloud GPU bills with an autoscaling deep learning cluster by @RiseML https://t.co/Lo8A3vGPkY #deeplearning #machinelearning #gpu @awscloud #tensorflow #autoscaling @kubernetesio

[heise Developer @heisedc]

RiseML bringt Deep Learning zu Kubernetes https://t.co/aN2d6H39Mr #Kubernetes #MachineLearning

[RiseML @riseml]

✨ Introducing RiseML — A platform for deep learning on Kubernetes https://t.co/hokv3qZq1d #deeplearning #kubernetes #machinelearning

[ECIR 2018 @ecir2018]

ECIR is now ranked A in the prestigious CORE ranking! https://t.co/i8SYBiggRd