Searching Relevant Variable Subsets in Complex Systems Using K-Means PSO

The Relevance Index method has been shown to be effective in identifying Relevant Sets in complex systems, i.e., variable sub-sets that exhibit a coordinated behavior, along with a clear independence from the remaining variables. The need for computing the Relevance Index for each possible variable sub-set makes such a computation unfeasible, as the size of the system increases. Because of this, smart search methods are needed to analyze large-size systems using such an approach. Niching metaheuristics provide an effective solution to this problem, as they join search capabilities to good exploration properties, which allow them to explore different regions of the search space in parallel and converge onto several local/global minima.
In this paper, we describe the application of a niching metaheuristic, K-means PSO, to a set of complex systems of different size, comparing, when possible, its results with the ground truth represented by the results of an exhaustive search, while we rely on the analysis of a domain expert to assess the results of larger systems. In all cases, we also compare the results of K-means PSO to another metaheuristic, based on a niching genetic algorithm, that we had previously developed.