AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Importance of Explainability in Machine Learning
The H2O stack was one of the first AutoML cases where you can just throw your same train test data set into a regression, random forest, neural net, and extra boost model to compare performance. But in the end, our job as data scientists, it's not to build a model, but to pick the right model and then build that. It's always an equation between accuracy on the one end and explainability on the other hand. And for that, we usually need simple methods rather than complex methods.