CS 2750 Machine Learning
Weak learnability=Strong (PAC) learnability
Assume there exists a
it is better that a random guess (50 %) with confidence
higher than 50 % on any data distribution
Is problem also PAC-learnable?
Can we generate an algorithm
that achieves an arbitrary
Why is important?
Usual classification methods (decision trees, neural nets),
have specified, but uncontrollable performances.
Can we improve performance to achieve pre-specified