Speaker 1: Qing Ji

Title

An Introduction to Markov Switching Models

Abstract

Markov switching models are a natural extension of the finite mixture models, which could be useful when error source are potentially from multiple different distributions and its label is missing.  I will give a short introduction on the basics of the finite mixture models and Markov chain, and combine the two to introduce Markov switching model and ways of estimation and inference.  Lastly, I will briefly explain my current work involving Fisher information matrix of Markov switching models.

Speaker 2: Maria Deliyianni

Title

Finite Element Method in 1 Dimension

Abstract

In this talk we will discuss about the Finite Element Method in approximating the solution to a boundary value problem in 1-D on the interval [a,b]. By converting the problem into the variational form, we will be able to discretize it and seek for an approximation of the solution in a finite dimensional subspace. This is obtained by partitioning the interval [a,b] and approximate the solution in each subinterval using basis functions of the finite dimensional subspace.