Hacker News Reader

Comment by sungam | Hacker News Reader

Yes you are right - the representation is biased due to the image dataset that I have used.

I don't think it would be useful to match the population distribution since the fraction of skin cancers would be tiny (less than 1:1000 of the images) so users would not learn what a skin cancer looks like, however in the next version I will make it closer to 50:50 and highlight the difference from the population distribution.

View on HN

Yes. As I said matching the population base rate wouldn't be practical, so you'd need to educate on that separately from the identification learning.

Let's say I achieve a 95% on the app though. Most people would have a massively over-inflated sense of their correctness in the wild. If the actual fraction is only 1/1000 and I see a friend with a lesion I identify as concerning, then my actual success rate would be:

    1*0.95 / (0.05*999 + 1*0.95)

So ~1.8%, not 95%. Few people understand Bayesian updating.

Thanks for this - I need to look at this more carefully

Reply on news.ycombinator.com