search

memoiry / LightML.jl

Minimal and clean examples of machine learning algorithms implemented in Julia
Other
404 stars 69 forks source link

normalize `err` #7

Closed szcf-weiya closed 6 years ago

szcf-weiya commented 7 years ago
  1. err need to be normalized. In AdaBoost algorithm, $err_m = \frac{\sum w_i I(y_i\neq G_m(x_i))}{\sum w_i}$;
  2. this might be a typo, 1 - 0.5 should be 1 - err.