ASI: Accuracy-Stability Index for Evaluating Deep Learning Models

Wei Dai, Daniel Berleant

In the context of deep learning research, where model introductions continually occur, the need for effective and efficient evaluation remains paramount. Existing methods often emp…