Penalized estimating equations for a stochastic model for compositional data

Omar De la Cruz Cabrera, Department of Mathematical Sciences, Kent State University, USA

We study several approaches for generating penalized estimating equations for the parameters of a stochastic process that models the evolution in continuous time of compositional data (i.e., vectors of positive numbers that add up to 1). The parameters to be estimated specify the Dirichlet distribution that is the invariant distribution of the process, as well as a time scale parameter. Regularization by penalization becomes especially important when the number of categories (and thus the number of parameters) is large, as is the case in many applications. This is joint work with Oana Mocioalca and Yicheng Su.

Please direct all questions to the following contacts - Your question will be routed to get you an answer as quickly as possible.

• MSUStatSymposium2018@stt.msu.edu
• Michigan State University
  Department of Statistics and Probability
  Room C413 Wells Hall
  619 Red Cedar Road
  East Lansing, MI 48824