Improving Generalization in Survival Models

Suggestions for estimating and enhancing predictive accuracy for the employee attrition case

Recently, I’ve come up with a particular issue when dealing with survival analysis: many models I fitted performed well in theory, with strong test metrics, but then failed to predict the true outcomes that were observed in practice. In this article, I want to discuss ways to better estimate the performance of our survival models, and a practical tip to help with …

accuracy analysis attrition employee employee retention improving israel issue metrics practice predictive suggestions survival survival-analysis test theory tips-and-tricks train-test-split true