Abstract:
Assessing risk in a computational grid environment is an essential need for a
user who runs applications from a remote machine on the grid, where resource
sharing is the main concern. As grid computing is the ultimate solution believed to
meet the ever-expanding computational needs of organizations, analysis of the
various possible risks to evaluate and develop solutions to resolve these risks is
needed. For correctly predicting the risk environment, we made a comparative
analysis of various machine learning modeling methods on a dataset of risk factors.
First we conducted a survey with International experts about the various risk factors
associated with grid computing. Second we assigned numerical ranges to each risk
factor based on a generic grid environment. We utilized data mining tools to pick the
contributing attributes that improve the quality of the risk assessment prediction
process. Finally,we modeled the prediction process of risk assessment in grid
computing utilizing Meta learning approaches in order to improve the performance
of the individual predictive models. Prediction of risk assessment is demanding
because it is one of the most important contributory factors towards grid computing.
Hence, researchers were motivated for developing and deploying grids on diverse
computers, which is responsible for spreading resources across administrative
domains so that resource sharing becomes effective. We present an adaptive neurofuzzy
inference system that can provide an insight of predicting the risk environment.
Also, we used a function approximation tool, namely, flexible neural tree for risk
prediction and risk (factors) identification. Flexible neural tree is a feed forward
neural network model, where network architecture was evolved like a tree. Our
comprehensive experiment finds score for each risk factor in grid computing together
with a general tree-based model for predicting risk.The empirical results illustrate
that the proposed framework is able to provide risk assessment with a good accuracy.
We concluded that data mining tools can provide further steps in building a risk
assessment model in a Grid environment with good accuracy, according to the
obtained empirical results.
