I am trying to solve the first one of these exercises:

enter image description here

The zero-one loss is 1 if a sample is misclassified and 0 if it classified correctly. The empirical risk is the sample mean of the loss. The true risk is the expectation of the loss.

I do not know how to go about this. I could conduct a one-sided t-test:

$H_0:$ The true risk is greater or equal to 0.25

$H_1:$ The true risk is smaller than 0.25

But the p-value, I learned, is not the probability of the Null-Hypothesis, so this will not help.

I also thought that the true risk is not a random variable? Is this question solvable?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.