On an integrated approach to member selection and parameter estimation for Pearson distributions (Q1061480): Difference between revisions

A sequential parameter selection and a fitting to a Pearson density distribution through special loss functions are discussed. Such loss functions can be the log-likelihood: ln f(x\({}_ i;\mu,\sigma^ 2,\sqrt{\beta_ 1},\beta_ 2)\). A numerical example is given. Analytical derivatives of Pearson density functions are provided.