Many people use the MSE decomposition to illustrate bias and variance. However, is there any statistical learning theory connecting these concepts? Namely, is there a formula calculating model capacity using bias and variance?
I think this issue is perfectly discussed and explained by Dr. Kilian Weinberger from Cornel University.
Here are the lecture notes: http://www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote12.html
You can also find video recordings of the course, so you can follow up.