Chebyshev's inequality in French

(statistics) The theorem that in any data sample with finite variance, the probability of any random variable X that lies k or more standard deviations away from the mean is no more than 1/k2, i.e. assuming mean μ and standard deviation σ, the probability is: