On Model Calibration

We suggest a technique that adds a simple post-processing function to your model to get class probabilities. The resulting model is well-calibrated. This improves model interpretability, is highly useful for the practical application of your model, and will increase the likelihood of regulatory success.

How Much Test Data Is Enough?

A very common and essential question when configuring a machine learning training process is how much test data we need such that the measured prediction performance can be trusted. Even though this is a very important question, it is too simplified.