In risk management systems the risk factors are described both in qualitative and
quantitative terms, therefore models have to be developed that can handle both input
types. Since these systems are full of uncertainty and qualitative description the
system parameters can be represented in linguistic form, using soft computing - for
example fuzzy-based - methods. These techniques are suitable for representing the
system outputs with more realistic and user-friendly result. The models analyzed in
the paper define the risk level of physical exercise-based fuzzy approximate reasoning
systems and structural risk management models. Clustering of risk factors and using
hierarchical decision structure in these models makes the evaluation simplified and
easy expandable. The models were analyzed in Matlab environment, using input data
selected from the theoretical parameters of several groups of patients. Based on this
test the novel theoretical model is suitable for comparison with previously implemented
models developed with algorithms by others in order to validate our model. The long-term
goal is the real-time risk assessment for the model in question while finding the
appropriate sensors and its systematization is in progress.
