Related Posts
What’s everyone up to today?
Additional Posts in Data & Analytics Consultants
Are there RPA use cases in data work?
What is a data lake in basic terms?
Thought this was interesting. Across 160 teams of researchers, just about all failed to make good life outcome predictions on things like GPA, evictions, layoffs, and others. Data followed 4.5k families across 15 years, with 13k features (varied over time). Haven't looked at it directly yet, but will be turning the docs and data inside out... In the meantime, authors claim this as showing the limits of ML. Oh, and it's published in PNAS, so you know there's some big publication energy there.
https://www.pnas.org/content/117/15/8398
New to Fishbowl?
unlock all discussions on Fishbowl.



CPUs can do practically anything but only a relatively few things at a time. GPUs are only good for very specific tasks, but can do hundreds or thousands of those things at a time. GPU is a kind of “math accelerator” for the computer.
Bowl Leader
GPUs are expensive parallelization components with non-negligible overhead. Need lots of data and/or very large models to justify their use. Hence their frequent use in DL and virtually nil use with other models, where CPU vectorization (which is crazy fast) does enough
Garden hose vs sprinkler: lots of throughput at one point vs less throughput but at many smaller points