Search Results for "requirement"

...In particular, we use an unsupervised formulation of the Random Forest algorithm to calculate similarities and provide them as input to a clustering algorithm. For the sake of efficiency and meeting the dynamism requirement of autonomic clouds, our methodology consists of two steps: (i) off-line clustering and (ii) on-line prediction. RF+PAM can: Cluster observations (Unsupervised Learning) Calculate the dissimilarity between 2 or more observations (how different two observations are)