The model is said to have a low bias if it abely predicts the training data labels. If the model makes too many mistakes on the training data, we say that it has a high bias, or that the model underfits the data.
#Reasons
- The model is too simple for the data (e.g. linear models often underfit)
- The features aren’t informative enough.
- You regularize too much.
#Possible Solutions
- Trying a more complex model
- Engineering features with higher Predictive Power
- Adding more training data when possible
- Reducing Regularization
Transformers