Biology Reference
In-Depth Information
between farms, Keeling et al. demonstrated the benefit of several control
strategies. A major effort in the development of large-scale ABM in epi-
demics is led by a collaborative group called MIDAS (Models of Infectious
Disease Agent Study).
3.5.2.3 Markov Chain Monte Carlo
An alternative to the deterministic SEIR model is a stochastic model using
a Markov chain Monte Carlo (MCMC) approach, which also incorporates
the biology of a disease. MCMC is a well-established method with a vast
range of applications; hence, there is immense literature available on this
method. (Readers are referred to Gilks et al. [2005] for thorough discussion
on MCMC.) Briefly, however, MCMC refers to a suite of methods that, in gen-
eral, draws samples from a target probability distribution. The states in an
MCMC model typically represent the disease stages, similar to the SEIR, and
possess Markov property wherein the condition of one's state is independent
of previous states and depends only on the current time. These states are
governed by transition probability.
MCMC is an attractive method for many reasons. It permits the flexible
modeling of variations in the infectious period over time (O'Neill 2002),
which cannot be modeled using SEIR framework. MCMC also allows the
model to reflect the stochastic nature of some disease transmission. Finally,
it avoids identification problems that are a large obstacle in applying the
more deterministic SEIR model. Morton and Finckenstadt (2005) describe an
example of how MCMC methods can be used in modeling infectious dis-
eases, such as measles incidence.
3.5.3 additional Methods
We have not yet exhausted all of the existing modeling frameworks, which
may be equally important, due to the scope of this topic. For example, geo-
statistical techniques that concern clustering, cluster detection, interpola-
tion, and disease mapping can certainly be used in conjunction with remote
sensing data and are very important to disease risk prediction. Readers are
referred to other publications for more information on such techniques (i.e.,
Elliott et al. 2000).
3.6 Conclusions
Remote sensing is an effective way to measure geophysical parameters impor-
tant to the transmission of infectious diseases. Risk assessments derived
from statistical or biological models using these parameters can reduce the
