simdat.mcgr: Simulated artificial data with covariates for 'MCGR'.
In senhu/mvClaim: Multivariate general insurance claims modelling

Description Usage Format Details Source

An artificial data set of size 1000 containing the bivariate response y1, y2, two covariates x1, x2, and the true label.

1	simdat.mcgr

A data frame with 1000 rows and 5 variables:

y1,y2: The bivariate response variable.
x1,x2: Two covariates.
label: Their true labels.

This data set of size 1000 is artificially generated as following: first x1 and x2 are simulated from Normal(0, 1). The true number of component is G = 2, and the two components are equally mixed, i.e. τ = (0.5, 0.5). Component 1 follows a Gumbel(2) copula, while component 2 follows a Frank(3) copula. For the marginal distributions, they all have gamma distributions, with shape parameters being γ1 = 6, γ2 = 3 for component 1, and γ1 = 2, γ2 = 2 for component 2. For component 1:

y1 = 1 + 0.1 x1 + 0.1 x2 ,

y2 = 2 + 0.2 x1 + 0.2 x2 .

For component 2:

y1 = 0.8 + 0.2 x1 + 0.2 x2 ,

y2 = 0.8 + 0.2 x1 + 0.2 x2 .

The final simulated data consists of 500 samples from component 1, 500 samples from component 2

Hu, S. and O'Hagan, A. (2019) Copula Averaging for Tail Dependence in General Insurance Claims Data. To appear.

senhu/mvClaim documentation built on Jan. 29, 2022, 3:18 p.m.