6.034 Tutorial: Adaboost

Pedro Cattori · 51:49

AdaBoost is presented as "adaptive" + "boosting" — take a bag of weak classifiers (ID trees, SVMs, neural nets, nearest neighbors), and combine them into H(x) = sign(Σ αᵢhᵢ(x)), adaptively re-weighting the training po...

Read the full summary on tuber

Redirecting...