A lock-gate sensor network keeps adding hidden layers on the same small labeled set, and validation error gets worse, not better. What is happening?
Select an answer to reveal the explanation.
Short Explanation
Think of a sensor network stacking hidden layers on a small labeled set, and validation error getting worse. Layer count is a capacity hyperparameter. Extra depth can overfit a small set.
Full Explanation
Network depth is a capacity hyperparameter. Adding hidden layers on a small labeled set can overfit and raise validation error. More layers do not always help. Rekognition is not that lever, and batch size is not the only way to overfit.