How do you handle outliers caused by heavy tails in an optimal way when you want to estimate (but not overestimate) the mean of a nonnegative random variable?
With Bart van Parys, we answer this question in our preprint
arxiv.org/abs/2503.21421
arxiv.org
Robust Mean Estimation for Optimization: The Impact of Heavy Tails
We consider the problem of constructing a least conservative estimator of the expected value $μ$ of a non-negative heavy-tailed random variable. We require that the probability of overestimating the e...
