Divide and Conquer: Recursive Likelihood Function Integration for Hidden Markov Models with Continuous Latent Variables

56 Pages Posted: 14 Jun 2016 Last revised: 11 Nov 2017

Date Written: November 7, 2017

Abstract

This paper develops a method to efficiently estimate hidden Markov models with continuous latent variables using maximum likelihood estimation. To evaluate the (marginal) likelihood function, I decompose the integral over the unobserved state variables into a series of lower dimensional integrals, and recursively approximate them using numerical quadrature and interpolation. I show that this procedure has very favorable numerical properties:

First, the computational complexity grows linearly in time, which makes the integration over hundreds and thousands of periods well feasible.

Second, I prove that the numerical error is accumulated sub-linearly over time; consequently, using highly efficient and fast converging numerical quadrature and interpolation methods for low and medium dimensions, such as Gaussian quadrature and Chebyshev polynomials, the numerical error can be well controlled even for very large numbers of periods.

Lastly, I show that the numerical convergence rates of the quadrature and interpolation methods are preserved up to a factor of at least 0.5 under appropriate assumptions.

I apply this method to the bus engine replacement model of Rust [Econometrica, 55 (5): 999–1033, (1987)]: first, I estimate the model using the original dataset; second, I verify the algorithm’s ability to recover the parameters in an extensive Monte Carlo study with simulated datasets.

Keywords: Hidden Markov Models, Numerical Quadrature, Interpolation, Recursion

JEL Classification: C25, C63

Suggested Citation

Reich, Gregor, Divide and Conquer: Recursive Likelihood Function Integration for Hidden Markov Models with Continuous Latent Variables (November 7, 2017). Available at SSRN: https://ssrn.com/abstract=2794884 or http://dx.doi.org/10.2139/ssrn.2794884

Gregor Reich (Contact Author)

Tsumcor Research AG ( email )

Switzerland

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
112
Abstract Views
781
Rank
522,828
PlumX Metrics