The discrepancy between a model’s projected values and observed or real values is known as bias. A model’s variance is the difference between the model’s predictions when fitted to the train and test data sets.
A model that is too basic will not be able to produce fully accurate predictions. The forecasts, on the other hand, will be constant. The model will be underfitting in this scenario, with strong bias and low variance. If the model is overly complicated, on the other hand, it might forecast accurately but not consistently. The model is considered to have high variance low bias in this scenario, implying that it will match the train data considerably better than the test data. An overfitted model is one that has been overfitted.