The expected convex hull trimmed regions of a sample (Q964641)

Let \(X=(x_1,\dots,x_n)\) be a sample of multivariate vectors. The author considers a set obtained by Minkovski averaging of convex hulls of all \(k\)-tupes \((x_{i_1},\dots, x_{i_k})\) from \(X\). This set is called the expected convex hull trimmed region of \(k\)-th order. Some properties of such regions (e.g., affine equivariance) and their volumes are established. An algorithm for their construction is described for bivariate data. Applications to financial mathematics are mentioned.