Eustasius Musenge

Short item Published on November 30, 2015

Modelling big “N” spatiotemporal data

In modelling hierarchical data we can take into account spatial and temporal correlations by introducing spatiotemporal random effects in the model. Several other hurdles have to be overcome when modelling hierarchical mortality data, but Bayesian techniques with the aid of the Markov chain Monte Carlo (MCMC) simulation methods have successfully overcome these and fit spatiotemporal random effects for reasonably sized geo-locations. However, as the number of geo-locations increases, MCMC computations become infeasible or extremely slow, which is a norm in Big Data Analytics (BDA). This problem is popularly known as the “big m” or “big N”.

Published on June 12, 2013 by

Rheumatoid arthritis disease progression in a South African cohort: Bayesian multistate chronic disease, dynamic modelling.

The primary goal of this article is to model the forces (rates) of recovery, relapse and mortality for patients started on rheumatoid arthritis (RA) standard treatment and the effect of adjusting for body functionality. A four state model inclusive of the absorbing state death was fit to these data. A consequence of an improved health is better body functionality, which was measured using a health assessment disability index (HAQ-DI). The modelling was done using a member of the generalised additive mixed models (GAMMs) which utilise nonparametric functions adjusting for over-dispersion and correlation. Based on the results we advocate that patients should be treated until the disease activity score is in remission or lowest possible to enable greater physical functionality whilst alleviating disability and mortality due to RA.