I have a machine that follows a probability distribution and it only spits out a value of 1 or -1.
If the mean and variance of this probability distribution are equal, is it more likely for the machine to spit out a value of 1 or a value of -1?
This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try
refreshing the page, (b) enabling javascript if it is disabled on your browser and,
finally, (c)
loading the
non-javascript version of this page
. We're sorry about the hassle.
Challenge Master Note: Wonderful interpretation of this problem! For completeness, you should show that there exists probabilities in interval [ 0 , 1 ] for these two outputs.
Log in to reply
Let P ( 1 ) = p . Then P ( − 1 ) = 1 − p and so
p ∗ ( 1 ) + ( 1 − p ) ∗ ( − 1 ) = μ ⟹ 2 p = μ + 1 ⟹ p = 2 μ + 1 = 4 5 + 1 ≈ 0 . 8 0 9 ,
and thus P ( − 1 ) = 1 − p = 4 3 − 5 ≈ 0 . 1 9 1 .
Problem Loading...
Note Loading...
Set Loading...
Variance, ( σ 2 ), is necessarily a non-negative value, and if both the mean, ( μ ), and variance were 0 then the only possible output would be 0 , which is not the case. Thus the variance, and thus also the mean, are positive, indicating that an output of 1 must be more likely than an output of − 1 .
More rigorously, given that σ 2 = μ , and noting that E ( X 2 ) = 1 since X 2 = 1 regardless of whether the output is − 1 or 1 , the formula σ 2 = E ( X 2 ) − μ 2 becomes
μ = 1 − μ 2 ⟹ μ 2 + μ − 1 = 0 ⟹ μ = 2 − 1 ± 5 .
As noted before, since variance is necessarily non-negative, we can conclude that μ = 2 5 − 1 = ϕ − 1 .