If I found X number of E events given Y unique samples of a known sample space S, where:
- Y is much smaller than is smaller than S
- E is a binary event
Can I say that event X has a probability and provide a confidence interval for that estimated probability?
For example, given a sample space of 10^4 black and white marbles in a bag. If I pulled 100 of those marbles and 10 of them were white, what could I determine about the probability of choosing a white marble out of the original/unsampled 10^4 bag of marbles?
This seems related to "Confidence / statistical significance of empirical probability estimation" but I know the sample space size.
