Probability that an external attacker infers the dominant contribution X1 to
within +/- beta (Proposition 5 of the paper):
mu_I(rho) = F_Z((1+beta)rho - 1) - F_Z((1-beta)rho - 1), with Z the
relative loss at dominance rho.
Usage
assess_risk_I(rho, sigma_nu, sigma_eps, n, beta)
Arguments
- rho
Dominance share X1/Y in (0,1]. Recycled.
- sigma_nu, sigma_eps, n
Mechanism parameters. Recycled.
- beta
Accuracy threshold of the inference.
Value
mu_I(rho), vectorised.
Examples
assess_risk_I(rho = 0.85, sigma_nu = 0.3, sigma_eps = 0.03, n = 6, beta = 0.2)
#> [1] 0.5647026