Solution strategy based on Gaussian mixture models and dispersion reduction for the capacitated centered clustering problem

The Capacitated Centered Clustering Problem (CCCP)—a multi-facility location model—is very important within the logistics and supply chain management fields due to its impact on industrial transportation and distribution. However, solving the CCCP is a challenging task due to its computational complexity. In this work, a strategy based on Gaussian mixture models (GMMs) and dispersion reduction is presented to obtain the most likely locations of facilities for sets of client points considering their distribution patterns. Experiments performed on large CCCP instances, and considering updated best-known solutions, led to estimate the performance of the GMMs approach, termed as Dispersion Reduction GMMs, with a mean error gap smaller than 2.6%. This result is more competitive when compared to Variable Neighborhood Search, Simulated Annealing, Genetic Algorithm and CKMeans and faster to achieve when compared to the best-known solutions obtained by Tabu-Search and Clustering Search.

Clustering web search results using Wikipedia resource

The paper presents a proposal of a new method for clustering search results. The method uses an external knowledge resource, which can be, for example, Wikipedia. Wikipedia – the largest encyclopedia, is a free and popular knowledge resource which is used to extract topics from short texts. Similarities between documents are calculated based on the similarities between these topics. After that, affinity propagation clustering algorithm is employed to cluster web search results. Proposed method is tested by AMBIENT dataset and evaluated within the experimental framework provided by a SemEval-2013 task. The paper also suggests new method to compare global performance of algorithms using multi – criteria analysis.

Feature Selection Clustering for Gene Data

Clustering is inherently a difficult task and is made even more difficult when the selection of relevant features is also an issue. In this paper , an algorithm is proposed which makes feature selection an integral part of the global clustering search procedure and attempts to overcome the problem of identifying less promising locally optimal solution in both clustering and feature selection. The proposed method uses genetic algorithm to preserve the population diversity and prevent premature convergence. The algorithm is implemented in Matlab 7.4 under windows operating system. The results show that the proposed algorithm outperforms existing algorithms in terms of accuracy.