More data is generally assumed to be good, and that having large amounts of data will generally result in better models with lower error rate. Which of the following are correct?
1) More data is always good
2) Having more data will help when the fitted model is underfitting
3) Having more data will help when the fitted model is overfitting
4) Having more data will help when the fitted model has high bias
5) Having more data will help when the fitted model has high variance