4 ms·
It's hard to imagine an narrow image classification task that humans will be able to beat NNs.
by JoshTko 7y ago
It's hard to imagine an narrow image classification task that humans will be able to beat NNs.
- hprotagonist 7y agoIt's not clear that melanoma diagnosis is fairly described as a static image classification task. Certainly it's a component, but (for example) patient history is, too.
- sgt101 7y agoConsidering that the ground truth comes from humans I would say that humans always outperform NN's and that results which show otherwise are demonstrating the limitations of the data set or testing process.
- leesec 7y agoIt's been proven that using humans as ground truth you can ultimately build a NN off that data which outperforms the humans.
- sgt101 7y agoWhat is meant by proven? What is meant by "the humans"? And how can any human say that is so?
- woadwarrior01 7y agoBecause an ensemble of humans can usually outperform a single human (quite like in artificial neural nets). One way to do it would be to setup a sufficiently large majority-voting ensemble of human classifiers.
- leesec 7y agoI mean, for example, you take 5 expert radiologists, and average their assessments when scoring an image, and your train a NN to predict those averaged scores, and with enough data the NN will beat any single selected doctor in accuracy.