3 ms·

I'm currently working on a machine learning project in healthcare and I'm kind of amazed by the lax attitude a lot of self-proclaimed data scientists seem to ha

I'm currently working on a machine learning project in healthcare and I'm kind of amazed by the lax attitude a lot of self-proclaimed data scientists seem to have about applying ML/DS methods to healthcare without involving any clinicians, because they insist that the "data doesn't lie"... there seems to be limited interest in exploring causality or the insights these methods provide. Instead the goal for so many is to transfer decision-making to models in the name of "efficiency". So many people in ML are haughty, arrogant hype-(wo)men, whose disinterest in the fields they are trying to 'disrupt' is gross. Please excuse the rant, but I'm so tired of this hype train. I agree with the author: people need to be aware of the limitations of machine learning models, but I'd add especially the people building them.

aix1 4y ago
I think this is a general problem with a lot of academic research. When someone develops a new technique for solving some abstract problem (a new graph algorithm, a new type of ML model etc) they want to demonstrate that it has practical applications. What often happens is that they take some problem domain they aren't experts on (their expertise is in maths, comp sci etc), apply their methods to the best of their knowledge and publish the results. Since such academics read each other's papers, this leads to a gradual divergence between what's in the literature from what's actually useful. I think one way to tackle this is by forming interdisciplinary teams. For example, I work at an industrial research lab on AI in healthcare, and our project team primarily consists of various clinical specialties. ML research and engineering are around 20% of the overall team.