Researchers find flaws in algorithm used to identify atypical medication orders
0
Enjoy being online again!
Welcome to the community of good people who base their values on evidence and appreciate civil discourse - the social network you will enjoy.Create your free account
2 comments
Feel free to reply to any comment by clicking the "Reply" button.Imagine a randomly selected person watching a doctor diagnose ten thousand patients.
When the next patient then comes in, the randomly selected person is asked to give a diagnosis. Having a good base of experience, this diagnosis might very well be correct, but ....
This is how an algorithm works
Complex algorithms can never be perfect, just as not all physicians are equally expert. They can disagree, as found in court litigation producing contrasting "expert witnesses".
Petter
Level 9
Nov 8, 2020
1
Knowing some of the limitations of AI, I cannot say that I am surprised.
anglophone
Level 9
Nov 8, 2020
2
Thirty five years ago I developed a "self-learning" algorithm to help our junior technicians identify problems in our company's control circuitry. It worked extremely well most of the time, and undoubtedly saved many hours of fault tracing time, but it was still, on occasion, stumped and a senior technician was required. Of course, each of these latter incidents helped the algorithm learn, but ....
@Petter Thanks for sharing that. 
Enjoy being online again!
Welcome to the community of good people who base their values on evidence and appreciate civil discourse - the social network you will enjoy.Create your free account
Share this post
Categories
Agnostic does not evaluate or guarantee the accuracy of any content. Read full disclaimer.