How to count m if the set of hypothesis is infinity?

- VC stated that eventhough the hypothesis space is infinite, as long as the sample is finite, then we can know how much the samples do we need.

- Proving lower bound is easier than proving an upper bound. Means that in this case we only need to to prove at least one example, than using max example
- For VC dimension, we need one example, but to prove that, test infinite number of hypothesis.
- In the vc dimension = 2, we know that we still get positive, in setting the limit [], between those two
- But in the case vc dimension = 3, one example exists that we can't use the VC, so therefore at least 2

- Here we have R2 dimension, and as an input, weight parameters wT*X and the output will be labeled by some hypothesis theta
- R2 will get 3 VC Dimensions.
- while R1 get 2VC, R2 get 3VC, does this get directly correlated by nVC = Rn-1?

- VC dimension directly correlated with the numbers of parameters.