Consider a variant of the Winnow algorithm which sets wit+1+3w, when it makes a mistake on positive examples (when Yt = 1 and ŷt = 0) and sets wit+1+wi,t/4 when it makes a mistake on negative examples (when y₁ = 0 and ŷ₁ = 1). Follow the same analysis steps as we did in the class to derive an upper bound for the number of mistakes of the algorithm for learning monotone disjunctions.

icon
Related questions
Question
100%
Consider a variant of the Winnow algorithm which sets wit+1+3w, when it makes a mistake on positive examples (when
Yt = 1 and ŷt = 0) and sets wit+1+wi,t/4 when it makes a mistake on negative examples (when y₁ = 0 and ŷ₁ = 1).
Follow the same analysis steps as we did in the class to derive an upper bound for the number of mistakes of the algorithm
for learning monotone disjunctions.
Transcribed Image Text:Consider a variant of the Winnow algorithm which sets wit+1+3w, when it makes a mistake on positive examples (when Yt = 1 and ŷt = 0) and sets wit+1+wi,t/4 when it makes a mistake on negative examples (when y₁ = 0 and ŷ₁ = 1). Follow the same analysis steps as we did in the class to derive an upper bound for the number of mistakes of the algorithm for learning monotone disjunctions.
Expert Solution
steps

Step by step

Solved in 2 steps

Blurred answer