AI glossary · Machine learning
Overfitting
Overfitting is when an AI model learns its training examples too closely, including their noise, so it does well on them but poorly on new data.
Also called underfitting
In more detail
An overfitted model has memorised rather than understood, like a student who learns past exam answers by heart but can’t handle a new question. The opposite problem, underfitting, is when a model is too simple to capture the real pattern.
Testing on data the model hasn’t seen, using more varied training data and data augmentation all help prevent it.
Related terms
Training dataTraining data is the set of examples, such as text, images or past records, that an AI model learns from. Its quality and range shape what the model can do.DatasetA dataset is an organised collection of data, such as a table of house prices or a folder of labelled photos, used to train or test AI models.Data augmentationData augmentation creates extra training examples by changing existing ones, such as flipping, cropping or recolouring photos, to help a model generalise.HyperparameterA hyperparameter is a setting chosen before training an AI model, such as how fast it learns or how many layers it has, that shapes how training goes.
Browse related tools
AI tools listed on BroUseAI in these categories.