Reproducibility in Machine Learning

The current state of reproduceability in a machine learning lig, is this a real problem? And you have any thoughts on how maybe we can solve it? A, i don't believe it's necessarily unique. It exists in many fields because the stakes of mistakes are relatively low. But if you say, up here is a thing, whatever i do, i beat everything else on the planet, ah, your paper has a very high likelihood of getting published,. less the reviewers really don't believe your experimental results.

Transcript

Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.