3 ms·

Counterpoint: regularization is just a way of specifying a Bayesian prior for maximum a posteriori estimation.

• mlevental 7y ago
  but what value is that perspective? how do i use this to actually fit a model?