Abstract:
Models involving the Kumaraswamy distribution have been a very studied in the past years
in the analysis and modeling of bounded continuos variables. In this paper we focus on one
in particular: the Trapezoidal Kumaraswamy model. We present an estimation method for
its parameters based on bayesian approaches: the Stochastic EM algorithm (SEM), which
avoids the most common issues of the classical EM. Then, we apply this method to the
daily covid-19 cases in Chile using this model.