Stability

Stability measures how much a learning algorithm’s output changes when its training data are altered. In statistical learning theory, it helps bound the gap between training performance and performance on new data.

Connect