Eight dominant Mediterranean species were classified into similar groups
according to their expected flammability, by applying multivariate statistical
methods (Hierarchical Cluster Analysis and Canonical Discriminant Analysis) on
the values of their most significant pyric properties (heat content, total and
mineral ash content, surface area-to-volume ratio, particle density). Based on
the statistical classification, meaningful explanations of the flammability
differences among individual species were deduced. The results were in good
agreement with similar rankings based on laboratory tests. Further validation
may render the method widely applicable for the assessment of species
potential flammability without laboratory flammability tests.
Real-time classification of IDS alerts with data mining techniques
