Tag: survey

Visual Analytics in Deep Learning: An Interrogative Survey for the Next Frontiers
By @fredhohman @minsukkahng @PoloChau

Bloghttps://t.co/2c1IUTYI8v
Paperhttps://t.co/acP2GVlE57 pic.twitter.com/OFmHxwjml9

— ML Review (@ml_review) October 18, 2018

I wrote an in-depth analysis of how GPUs would compare against TPUs for training BERT. I conclude that current GPUs are about 30-50% slower than TPUs for this task https://t.co/BG8mIqQWMj

— Tim Dettmers (@Tim_Dettmers) October 17, 2018

I've spent most of 2018 training models that could barely fit 1-4 samples/GPU.
But SGD usually needs more than few samples/batch for decent results.
I wrote a post gathering practical tips I use, from simple tricks to multi-GPU code & distributed setups: https://t.co/oLe6JlxcVw pic.twitter.com/pQTXQ9X7Ug

— Thomas Wolf (@Thom_Wolf) October 15, 2018

Nice post exploring chart themes in Altair. Themes are a feature we built into the package, but haven't done much with yet... great to see users running with it!https://t.co/A852tFkTM6

— Jake VanderPlas (@jakevdp) October 12, 2018

A Review of the Neural History of Natural Language Processing-- great neural NLP overview by @seb_ruder https://t.co/NmicL0THJU

— Rachel Thomas (@math_rachel) October 11, 2018

Deep Learning with Electronic Health Record (EHR) Systems
By @GokuMohandashttps://t.co/V4ErxodSnz pic.twitter.com/CE34xzwTyu

— ML Review (@ml_review) October 10, 2018

I wrote a blog post about a bunch of ways to generate/plot uncertainty estimates https://t.co/OdumNDpQtS

— Erik Bernhardsson (@fulhack) October 10, 2018

Pipenv: The Future of Python Dependency Management by Kenneth Reitz - https://t.co/COWMcKjNru. This talk covers the history of Python packaging, the tools that have been historically available for app deployment, the problems/constraints presented by them, and more.

— Python Software (@ThePSF) October 7, 2018

👩‍💻 this really is such a great guide — even the summary list alone:
📄 "Good enough practices in scientific computing" by @gvwilson, @JennyBryan, @kcranstn, @lexnederbragt, @tracyktealhttps://t.co/LXXP1IY99c

— Mara Averick (@dataandme) October 7, 2018

After 3 years, I think we are converging on a solid course to introduce our @UBCMDS students to the Data Science tools/software stack (e.g., Jupyter, Git & @github, @rstudio , Rmd, etc) and are now ready to share the materials: https://t.co/wSqrKkU5Se

— Tiffany Timbers (@TiffanyTimbers) October 1, 2018

A (Long) Peek into Reinforcement Learning
By @lilianwenghttps://t.co/NpKaSpFQ88 pic.twitter.com/aaT7PPYcFU

— ML Review (@ml_review) September 9, 2018

Those interested in forecasting you can look at the new blog by Slawek Smyl of Uber whose Hybrid method produced the most accurate forecasts in the M4 Competition. It provides an excellent introduction to business forecasting.
https://t.co/Eo6feY6jp3@SlawekSmyl

— Spyros Makridakis (@spyrosmakrid) September 8, 2018