1
$\begingroup$

This question already has an answer here:

Is it correct to say that, in the Bias-Variance trade-off:

the bias error represents the ability of the model to possibly map the trend of the training set ?

and the variance represents how much the model is prone to variations if the dataset changes ?

$\endgroup$

marked as duplicate by Stephen Rauch Oct 15 at 13:06

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

1
$\begingroup$

I would say that both the definitions you gave above refer to variance.

Variance is an index of model complexity. Thus:

high variance -> high complexity -> many degrees of freedom in model -> it overfits training data -> if dataset changes, accuracy degrades a lot

Bias is an index of the systematic error of the model, as it shows the difference between the expected value of various model's predictions for a given target output and the ground truth.

$\endgroup$
0
$\begingroup$

Bias are the simplifying assumptions made by a model to make the target function easier to learn.

Variance is the amount that the estimate of the target function will change if different training data was used.

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.