The projection-based multivariate density estimation
DOI:
https://doi.org/10.12697/ACUTM.2004.08.09Keywords:
multivariate distribution density, density estimation, projection pursuit, Gaussian distribution densityAbstract
The paper discusses methods of estimation of the multivariate density function by using statistical estimates of the densities of the univariate sample projections. The Gaussian mixture model is analyzed. Density parametrization is applied and parameter calculation methods based on the parameters of projections are proposed. A simulation study is carried out and the simulation results are discussed.
Downloads
Download data is not yet available.