Prove that for every learning algorithm A there exist a probability distribution, D, and a learning algorithm B such that A is not better than B w.r.t. D. January 19, 2021 Read More »
Show that if H is PAC learnable (in the standard one-oracle model), then H is PAC learnable in the two-oracle model. January 19, 2021 Read More »
Prove that Equation (5.2) suffices for showing that P[LD(A(S))≥ 1/8] ≥ 1/7. January 19, 2021 Read More »
Show that there are cases in which the previous two inequalities are strict (namely, the ≤ can be replaced by ) and cases in which they can be replaced by equalities. January 19, 2021 Read More »
Use the Pigeonhole principle to show that there must be a pair i j ≤ d + 1 such that _i and _ j use the same xk and use that fact to derive a contradiction to the requirements from the conjunctions hi ,h j . January 19, 2021 Read More »
Prove that for every H that is PAC learnable, VCdim(H)∞. (Note that this is the implication 3→6 in Theorem 6.7.) January 19, 2021 Read More »
Show that for every g : Rn →R and every vector space of functions F as defined earlier, VCdim(POS(F +g))= VCdim(POS(F)). January 19, 2021 Read More »
Show that each of the following classes can be represented as a Dudley class: January 19, 2021 Read More »
Prove that the class of all polynomial classifiers over R has infinite VCdimension. January 19, 2021 Read More »