Hello,

You won't be able to have the exact probability...

You can have an upper bound for the probability with Chebyshev's inequality.

And you can have an approximation thanks to the CLT.

In both cases, it would depend on , I guess that's what is missing.