Box 1.
Common definitions.
Transfer learning: Applying the knowledge gained from one domain to another one, for example, using a model trained using images from one cell line (source domain) to predict images from another cell line (target domain) Label smoothing: Acknowledging possible errors in annotations and setting the prediction target to values away from traditional 0/1 (e.g., 0.1 and 0.9) to reflect this.42 Evaluation metrics: ![]() |