0100
A good model should capture valuable patterns in the data and discard any noise that doesn't help with predictions. An overfitting model will fit that noise. An underfitting model will not capture the relevant patterns in the dataset. + +An overfitting model should not have any problems with the training data, so we should expect a low training loss. An underfitting model should struggle with the training data, so its training loss will be high. + +This model shows a high training loss, which we expect for an underfitting model.
* Check ["Overfitting and Underfitting with Learning Curves"](https://articles.bnomial.com/overfitting-underfitting-learning-curves) for an introduction to two fundamental concepts in machine learning through the lens of learning curves.