Discover and read the best of Twitter Threads about #ml4h
Most recents (2)
1/NEW: Today the SPIRIT-AI and CONSORT-AI Working Group have published the first reporting guidelines for clinical trials evaluating AI interventions in @NatureMedicine @LancetDigitalH @bmj_latest 
2/Papers here:
SPIRIT-AI nature.com/articles/s4159…; doi.org/10.1016/S2589-…; bmj.com/content/370/bm…
CONSORT-AI nature.com/articles/s4159…; doi.org/10.1016/S2589-…; bmj.com/content/370/bm…
SPIRIT-AI nature.com/articles/s4159…; doi.org/10.1016/S2589-…; bmj.com/content/370/bm…
CONSORT-AI nature.com/articles/s4159…; doi.org/10.1016/S2589-…; bmj.com/content/370/bm…
3/We need robust evaluations of *effectiveness* for AI interventions. As we move towards RCTs, #spiritAI and #consortAI will be crucial to ensuring robust, quality evidence.
A few nuggets from @geoffreyhinton's talk from earlier today at the #ml4h unconference. First up, the distinction between statistics and AI (and presumably ML by implication). Overall, I think these are pretty clean contrasts:
On interpretability: It's going to be very hard to explain the contribution of 1000s of statistical regularities that a net uses to produce a decision. Humans are really bad at explaining their decisions (most are post-hoc), so we shouldn't hold neural nets to a double standard
Related to interpretability: Regulating neural nets should be based on how it performs, not based on how it works. This is what we do with people too.
