Statistical Evaluation of Aggregates
A variety of statistical techniques can be used to evaluate the durability of aggregate. Most of these involve multivariate statistical techniques, such as K-means clustering, Discriminant analysis, Multivariate regression analysis, Tree clustering, and Factor Analysis.
Using these analyses in combination, it is possible to classify aggregate into groups of similar properties, quality, and durability by considering their service record and laboratory test results. The techniques are rather specialized, but the results are easily interpreted.
An example of such analyses is given in following papers:
QUANTITATIVE PETROGRAPHIC EVALUATION OF FINE AGGREGATE - use of multivariate regression analysis to obtain equations of contribution of various tests to quality index for fine aggregates.
STRESS RELAXATION OF ROCKS AND ITS RELATIONSHIP TO ROCK DURABILITY PROPERTIES - use of K-means clustering, Discriminant, and Factor analyses.
