Quote:
Originally Posted by yaser
There are two factors here: the level of deterministic noise, and the complexity of the hypothesis set used to fit the data. Both of these factors are affected by having the target fixed while expanding the hypothesis set. The two effects push overfitting in different directions.
|
Thanks again Professor, I have just one more question.
The two effects push overfitting in different directions, this means that if we fix a target, and expand H, the more complex H has a better ability to tackle the deterministic noise, but the complexity of the H in turn, makes for bad overfitting?