Alexandre Szabo

The proposal of a fuzzy clustering algorithm based on particle swarm

This paper proposes the Fuzzy Particle Swarm Clustering (FPSC) algorithm, which is an extension of the crisp data clustering algorithm PSC particularly tailored to deal with fuzzy clusters. The main structural changes of the original PSC algorithm to design FPSC occurred in the selection and evaluation steps of the winner particle, comparing the degree of membership of each object from the database in relation to the particles in the swarm. The FPSC algorithm was applied to eight databases from the literature with the purpose of benchmarking and its performance was compared with that of Fuzzy C-Means and Fuzzy PSO. The results showed that the FPSC algorithm is competitive with the algorithms discussed in this paper.

A bee-inspired algorithm for optimal data clustering

The amount of data generated in different knowledge areas has made it necessary the use of data mining tools capable of automatically analyzing and extracting knowledge from datasets. Clustering is one of the most important tasks in data mining and can be defined as the process of partitioning objects into groups or clusters, such that objects in the same group are more similar to one another than to objects belonging to other groups. In this context, this paper aims to propose an adaptation of a bee-inspired optimization algorithm so that it is able to solve data clustering problems. The algorithm was run for different datasets and the results obtained showed high quality clusters and diversity of solutions, whilst a suitable number of clusters was automatically determined.

FaiNet: An immune algorithm for fuzzy clustering

Data clustering is useful in several areas, such as web mining, biology, climate, medical diagnosis, computer vision, marketing and others. Thus, in real problems, data can simultaneously belong to more than one cluster, being necessary to use fuzzy clustering concepts as decision mechanisms to assign data into clusters. Moreover, nature-based intelligent mechanisms have been used to increase the effectiveness of several machine learning algorithms. This paper proposes improvements on aiNet (Artificial Immune Network), a bioinspired clustering algorithm, and its extension to be applied to fuzzy partitions. The modified algorithm to be applied in fuzzy partitions was thus named FaiNet (Fuzzy aiNet). It uses immune system concepts to allow it to automatically detect a suitable number of clusters in the datasets, what is not possible for most clustering algorithms. FaiNet was applied to seven databases from the literature with the purpose of benchmarking and its performance was compared with that of Fuzzy C-Means, a Fuzzy particle swarm clustering algorithm (FPSC) and the improved crisp aiNet. Purity and Entropy were the main metrics used to evaluate performance. The FaiNet algorithm showed to be competitive with the other algorithms used for comparison.

The Grand Challenges in Natural Computing Research: The Quest for a New Science

An important premise of Natural Computing is that some form of computation goes on in Nature, and that computing capability has to be understood, modeled, abstracted, and used for different objectives and in different contexts. Therefore, it is necessary to propose a new language capable of describing and allowing the comprehension of natural systems as a union of computing phenomena, bringing an information processing perspective to Nature. To develop this new language and convert Natural Computing into a new science it is imperative to overcome three specific Grand Challenges in Natural Computing Research: Transforming Natural Computing into a Transdisciplinary Discipline, Unveiling and Harnessing Information Processing in Natural Systems, Engineering Natural Computing Systems.

Particle Swarm Clustering in clustering ensembles

Graphical abstractDisplay Omitted HighlightsA new consensus function based on the Particle Swarm Clustering algorithm.An alignment-free efficient representation for both disjoint and overlapping partitions.Employment of evolutionary operators for ensemble pruning. A clustering ensemble combines in a consensus function the partitions generated by a set of independent base clusterers. In this study both the employment of particle swarm clustering (PSC) and ensemble pruning (i.e., selective reduction of base partitions) using evolutionary techniques in the design of the consensus function is investigated. In the proposed ensemble, PSC plays two roles. First, it is used as a base clusterer. Second, it is employed in the consensus function; arguably the most challenging element of the ensemble. The proposed consensus function exploits a representation for the base partitions that makes cluster alignment unnecessary, allows for the combination of partitions with different number of clusters, and supports both disjoint and overlapping (fuzzy, probabilistic, and possibilistic) partitions. Results on both synthetic and real-world data sets show that the proposed ensemble can produce statistically significant better partitions, in terms of the validity indices used, than the best base partition available in the ensemble. In general, a small number of selected base partitions (below 20% of the total) yields the best results. Moreover, results produced by the proposed ensemble compare favorably to those of state-of-the-art clustering algorithms, and specially to swarm based clustering ensemble algorithms.

A constructive particle swarm algorithm for fuzzy clustering

This paper proposes a fuzzy version of the crisp cPSC (Constructive Particle Swarm Clustering), called FcPSC (Fuzzy Constructive Particle Swarm Clustering). In addition to detecting fuzzy clusters, the proposed algorithm dynamically determines a suitable number of clusters in the datasets without the need of prior knowledge, necessary in cPSC to control the number of particles in the swarm. The FcPSC algorithm was applied to six databases from the literature and its performance was compared with that of Fuzzy C-Means, a Fuzzy Artificial Immune Network, a Fuzzy Particle Swarm Clustering and the crisp cPSC. FcPSC showed to be competitive with the algorithms used for comparison and the number of particles generated was smaller than for cPSC.

cPSCLASS: A CONSTRUCTIVE PARTICLE SWARM CLASSIFIER

The data classification task is one of the main tasks within the knowledge discovering from databases field. Its goal is to allow the correct classification of new objects (records from a database), unknown to the classifier, based upon the extraction of knowledge from objects whose classes are known a priori. The known data can be used to generate a classification model, or simply to infer the class of new objects from those whose classes are known. This paper presents a proposal for a classification algorithm, called Constructive Particle Swarm Classifier (cPSClass), which uses mechanisms from the Particles Swarm Clustering algorithm and Artificial Immune Systems to determine dynamically the number of prototypes from a database and use them to predict the correct class to which a new input object should belong. For performance evaluation the cPSClass was applied to several datasets from the literature and its performance was compared with that of its predecessor version, the nonconstructive Particle Swarm Classifier, and also to some classic algorithms from the literature.

The proposal of a Constructive Particle Swarm Classifier

The data classification task is one of the main tasks within the knowledge discovering from databases (KDD). Its goal is to allow the correct classification of new objects (records from a database), unknown to the classifier, based upon the extraction of knowledge from objects known a priori. These data already known can be used to generate a classification model, or simply to infer the class of new objects, from those whose classes are known. This paper presents a proposal for a classification algorithm, called Constructive Particle Swarm Classifier (cPSClass), which uses mechanisms from the Particles Swarm Clustering algorithm and Artificial Immune Systems to determine dynamically the number of prototypes from a database and use them to predict the correct class to which a new input object should belong. For performance evaluation the cPSClass was applied to some datasets from the literature and its performance was compared with its predecessor version, the non constructive Particle Swarm Classifier, and also the Naïve Bayes algorithm.

A Particle Swarm Clustering Algorithm with Fuzzy Weighted Step Sizes

This paper proposes a modification in the Fuzzy Particle Swarm Clustering (FPSC) algorithm such that membership degrees are used to weight the step size in the direction of the local and global best particles, and in its movement in the direction of the input data at every iteration. This results in the so-called Membership Weighted Fuzzy Particle Swarm Clustering (MWFPSC). The modified algorithm was applied to six benchmark datasets from the literature and its results compared to that of the standard FPSC and FCM algorithms. By introducing these modifications it could be observed a gain in accuracy, representativeness of the clusters found and the final Xie-Beni index, at the expense of a slight increase in the practical computational time of the algorithm.