9 Predict Future Climate Change Using Artificial Neural Networks
59
9.3 Related Work
An investigation was carried out for the problem of ensemble learning in raster
classification. This problem is important in many applications, such as classification
in medical image processing and land cover classification in remote sensing. The
problem is challenging due to the effect of class ambiguity from spatial heterogeneity.
The artificial neural network approach can be used to solve nonlinear problems
of regression that arise in environmental modeling, which includes forecasting for
a short term period in addition to rainfall-runoff modeling and atmospheric concentrations that leads to pollution. The aim was to review the existing methodology for
estimating predictive uncertainty as environmental datasets are redundant and noisy,
which describes how a predictive distribution model can be used to assess the impacts
occurred due to the climate change and to improve the different decision to avoid the
different impacts of it (Cawley et al. 2007).
An investigation was carried out to determine the relationship between different
metrological factors and Newcastle disease and the different key factors that cause
this disease were determined. To achieve this, a Newcastle disease forecasting model
was built and a Back Propagation neural network classification technique was applied
for animal disease forecasting in their research. For discovery covariate which is
based on a Poisson description with high-frequency Bayesian framework and Sparse
regression model of hierarchical Bayesian was used to identify the covariates that
affect the precipitation frequency which are collected from the various observations
at different stations over many different climatologically regions which are in the
U.S. continental (Debasish Das et al. 2014).
9.4 Artificial Neural Network
Artificial neural networks first learn by collecting the obtained data in the system, then
obtain results using the information they have learned against the samples that are
never presented to the network. Because of these learning and generalization features,
artificial neural networks have found a wide range of application opportunities in
many fields of science and have demonstrated the ability to solve complex problems
successfully. According to another definition, artificial neural networks are parallel
and distributed information processing structures that are inspired by the human
brain and are connected to each other by means of predominant connections where
each of which has its own memory; in other words, they are computer programs that
mimic biological neural networks.
59
9.3 Related Work
An investigation was carried out for the problem of ensemble learning in raster
classification. This problem is important in many applications, such as classification
in medical image processing and land cover classification in remote sensing. The
problem is challenging due to the effect of class ambiguity from spatial heterogeneity.
The artificial neural network approach can be used to solve nonlinear problems
of regression that arise in environmental modeling, which includes forecasting for
a short term period in addition to rainfall-runoff modeling and atmospheric concentrations that leads to pollution. The aim was to review the existing methodology for
estimating predictive uncertainty as environmental datasets are redundant and noisy,
which describes how a predictive distribution model can be used to assess the impacts
occurred due to the climate change and to improve the different decision to avoid the
different impacts of it (Cawley et al. 2007).
An investigation was carried out to determine the relationship between different
metrological factors and Newcastle disease and the different key factors that cause
this disease were determined. To achieve this, a Newcastle disease forecasting model
was built and a Back Propagation neural network classification technique was applied
for animal disease forecasting in their research. For discovery covariate which is
based on a Poisson description with high-frequency Bayesian framework and Sparse
regression model of hierarchical Bayesian was used to identify the covariates that
affect the precipitation frequency which are collected from the various observations
at different stations over many different climatologically regions which are in the
U.S. continental (Debasish Das et al. 2014).
9.4 Artificial Neural Network
Artificial neural networks first learn by collecting the obtained data in the system, then
obtain results using the information they have learned against the samples that are
never presented to the network. Because of these learning and generalization features,
artificial neural networks have found a wide range of application opportunities in
many fields of science and have demonstrated the ability to solve complex problems
successfully. According to another definition, artificial neural networks are parallel
and distributed information processing structures that are inspired by the human
brain and are connected to each other by means of predominant connections where
each of which has its own memory; in other words, they are computer programs that
mimic biological neural networks.