News for hackers. Stuff that matters.

A comment on Building an AI to predict human age from a blood sample. Parent comment

Re:Building an AI to predict human age from a blood sample by fock on Monday March 09, 2020 @04:31AM (#22523880)
that's exactly the thing I'm seeing in my field (computational materials science). Basically a simple regression model (with very simple features) brings you 90% there, still people compete on publishing (on ONE shitty benchmark dataset) ever better results – the most-cited people are using KRR, e.g. each fit uses 2TB of RAM and "days" of CPU (features of length O(1000) and 100000 samples). The sample data is probably a 30s calculation (and still only a rough estimation). Sometimes wants you to question science, but hey, writing proposals with "ML" in it gives at least a chance on that grant...
[ Parent | on HN ]
Re:Building an AI to predict human age from a blood sample by nilkn on Monday March 09, 2020 @11:16AM (#22526040)
This is great if you only need 90%. But often that extra 10% is the difference between a state of the art commercially viable product and vaporware that nobody will pay for.

It’s sort of like production software in general. Sometimes the core product is pretty easy to prototype... but serving it to all your users with very high reliability and uptime is not so easy, and that’s what actually gets them to pull out the credit card.

[ Parent | on HN ]
Re:Building an AI to predict human age from a blood sample by KaoruAoiShiho on Monday March 09, 2020 @07:06AM (#22524427)
I haven't thought as much about it as you have probably, but ok you get 90% of the way there, but now what? How do you get that last 10%? It would be a huge amount of work right?
[ Parent | on HN ]
Re:Building an AI to predict human age from a blood sample by hef19898 on Monday March 09, 2020 @09:17AM (#22525085)
Isn't that usually the case? That the first 90% are as difficult to achieve as the next 5%, which are again as hard as the next 3%?
[ Parent | on HN ]
Re:Building an AI to predict human age from a blood sample by KaoruAoiShiho on Monday March 09, 2020 @09:37AM (#22525218)
Basically the argument is that ML would be able to get there easier.
[ Parent | on HN ]
Re:Building an AI to predict human age from a blood sample by fock on Tuesday March 10, 2020 @04:18AM (#22533675)
you get the last 10% by running a calculation for that (which itself is not necessarily accurate...). The whole situation is a little bit like Plato's cave... And also from what I've seen so far, those models either are 90% there or they are overfitting like hell.
[ Parent | on HN ]