Multiobjective optimization for classifier ensemble and feature selection: an application to named entity recognition

In this paper, the concept of finding an appropriate classifier ensemble for named entity recognition is posed as a multiobjective optimization (MOO) problem. Our underlying assumption is that instead of searching for the best-fitting feature set for a particular classifier, ensembling of several classifiers those are trained using different feature representations could be a more fruitful approach, but it is crucial to determine the appropriate subset of classifiers that are most suitable for the ensemble. We use three heterogenous classifiers namely maximum entropy, conditional random field, and support vector machine in order to build a number of models depending upon the various representations of the available features. The proposed MOO-based ensemble technique is evaluated for three resource-constrained languages, namely Bengali, Hindi, and Telugu. Evaluation results yield the recall, precision, and F-measure values of 92.21, 92.72, and 92.46%, respectively, for Bengali; 97.07, 89.63, and 93.20%, respectively, for Hindi; and 80.79, 93.18, and 86.54%, respectively, for Telugu. We also evaluate our proposed technique with the CoNLL-2003 shared task English data sets that yield the recall, precision, and F-measure values of 89.72, 89.84, and 89.78%, respectively. Experimental results show that the classifier ensemble identified by our proposed MOO-based approach outperforms all the individual classifiers, two different conventional baseline ensembles, and the classifier ensemble identified by a single objective?Cbased approach. In a part of the paper, we formulate the problem of feature selection in any classifier under the MOO framework and show that our proposed classifier ensemble attains superior performance to it.     

A symmetry based multiobjective clustering technique for automatic evolution of clusters

In this paper the problem of automatic clustering a data set is posed as solving a multiobjective optimization (MOO) problem, optimizing a set of cluster validity indices simultaneously. The proposed multiobjective clustering technique utilizes a recently developed simulated annealing based multiobjective optimization method as the underlying optimization strategy. Here variable number of cluster centers is encoded in the string. The number of clusters present in different strings varies over a range. The points are assigned to different clusters based on the newly developed point symmetry based distance rather than the existing Euclidean distance. Two cluster validity indices, one based on the Euclidean distance, XB-index, and another recently developed point symmetry distance based cluster validity index, Sym-index, are optimized simultaneously in order to determine the appropriate number of clusters present in a data set. Thus the proposed clustering technique is able to detect both the proper number of clusters and the appropriate partitioning from data sets either having hyperspherical clusters or having point symmetric clusters. A new semi-supervised method is also proposed in the present paper to select a single solution from the final Pareto optimal front of the proposed multiobjective clustering technique. The efficacy of the proposed algorithm is shown for seven artificial data sets and six real-life data sets of varying complexities. Results are also compared with those obtained by another multiobjective clustering technique, MOCK, two single objective genetic algorithm based automatic clustering techniques, VGAPS clustering and GCUK clustering.     

Automatic MR brain image segmentation using a multiseed based multiobjective clustering approach

In this paper, the automatic segmentation of a multispectral magnetic resonance image of the brain is posed as a clustering problem in the intensity space. The automatic clustering problem is thereafter modelled as solving a multiobjective optimization (MOO) problem, optimizing a set of cluster validity indices simultaneously. A multiobjective clustering technique, named MCMOClust, is used to solve this problem. MCMOClust utilizes a recently developed simulated annealing based multiobjective optimization method as the underlying optimization strategy. Each cluster is divided into several small hyperspherical subclusters and the centers of all these small sub-clusters are encoded in a string to represent the whole clustering. For assigning points to different clusters, these local sub-clusters are considered individually. For the purpose of objective function evaluation, these sub-clusters are merged appropriately to form a variable number of global clusters. Two cluster validity indices, one based on the Euclidean distance, XB-index, and another recently developed point symmetry distance based cluster validity index, Sym-index, are optimized simultaneously to automatically evolve the appropriate number of clusters present in MR brain images. A semi-supervised method is used to select a single solution from the final Pareto optimal front of MCMOClust. The present method is applied on several simulated T1-weighted, T2-weighted and proton density normal and MS lesion magnetic resonance brain images. Superiority of the present method over Fuzzy C-means, Expectation Maximization clustering algorithms and a newly developed symmetry based fuzzy genetic clustering technique (Fuzzy-VGAPS), are demonstrated quantitatively. The automatic segmentation obtained by multiseed based multiobjective clustering technique (MCMOClust) is also compared with the available ground truth information.     

Template-free synthesis of flower-shaped zero-valent iron nanoparticle: Role of hydroxyl group in controlling morphology and nitrate reduction

Well-dispersed single phasic flower-like zero valent iron nanoparticles have been synthesized under aerobic conditions using a facile approach without the addition of any additives or templates. The role of hydroxyl groups of polyhydroxy alcohols in controlling surface morphology of nanoparticles has been thoroughly investigated. The obtained nanoparticles have been characterized by TEM, FE-SEM, XRD and BET surface area analyzer. Electron microscopy analyses reveal that the solvent plays a pivotal role in determining the morphology of the particles. With increase in viscosity of the solvent, formations of ‘petal-like’ structures, which are joined at the center are formed. The nitrate removal efficiency of the iron nanoparticles synthesized in different solvents has been studied and it is seen that the “flower-like” iron nanoparticles were most active in the removal of nitrate. Experiments have been done by varying (i) nitrate concentrations, (ii) nanoparticle dose, and (iii) type of nanoparticles. The results conclude that highest removal efficiency (～100%) was achieved when the nanoparticle dose was 2.88 g/L, even for high nitrate concentrations up to 400 mg/L. The major highlight of this work is the fact that even though the nanoparticles synthesized in glycerol-water mixture have larger size in comparison to the other nanoparticles, still they remove the nitrates with highest efficiency.”     

MAkE: Multiobjective algorithm for k-way equipartitioning of a point set

The classical problem of partitioning a given set of points, has applications in several areas such as facility location, scattered network, and in hierarchical design of VLSI circuits. While equipartitioning is traditionally associated with the single objective of minimum cutcost, the above application areas appear to demand more. In this paper, we introduce the problem of multiobjective k-way equipartitioning of a point set. Brief discussions on the above applications are followed by their generic formulation as a multiobjective k-way equipartitioning problem of a given point set. The non-commensurate multiobjective criteria addressed include (i) minimizing overall areas of the partitions, (ii) maximizing area of the individual partitions, (iii) minimizing the total compactness of the partitions, and (iv) minimizing the total geometric diversity of the obtained partitions. Since this optimization problem is computationally expensive in time and space, a technique based on genetic algorithm is proposed in order to obtain high quality results. Crossover and mutation operators specific to the k-way equipartitioning problem, have been designed and a new greedy operator named compaction is proposed to accelerate convergence. To illustrate the utility of the proposed formulation and the algorithm, a problem in VLSI layout design is considered. Results on synthetic data sets as well as those extracted from layouts of benchmark circuits demonstrate the effectiveness of the proposed multiobjective approach.     

Microblog summarization using self-adaptive multi-objective binary differential evolution

Social media platforms become paramount for gathering relevant information during the occurrence of any natural disaster. Twitter has emerged as a platform which is heavily used for the purpose of communication during disaster events. Therefore, it becomes necessary to design a technique which can summarize the relevant tweets and thus, can help in the decision-making process of disaster management authority. In this paper, the problem of summarizing the relevant tweets is posed as an optimization problem where a subset of tweets is selected using the search capability of multi-objective binary differential evolution (MOBDE) by optimizing different perspectives of the summary. MOBDE deals with a set of solutions in its population, and each solution encodes a subset of tweets. Three versions of the proposed approach, namely, MOOTS1, MOOTS2, and MOOTS3, are developed in this paper. They differ in the way of working and the adaptive selection of parameters. Recently developed self-organizing map based genetic operator is explored in the optimization process. Two measures capturing the similarity/dissimilarity between tweets, word mover distance and BM25 are explored in the optimization process. The proposed approaches are evaluated on four datasets related to disaster events containing only relevant tweets. It has been observed that all versions of the developed MOBDE framework outperform the state-of-the-art (SOA) techniques. In terms of improvements, our best-proposed approach (MOOST3) improves by 8.5% and 3.1% in terms of ROUGE??2 and ROUGE?L, respectively, over the existing techniques and these improvements are further validated using statistical significance t-test.

     

GAPS: A clustering method using a new point symmetry-based distance measure

In this paper, an evolutionary clustering technique is described that uses a new point symmetry-based distance measure. The algorithm is therefore able to detect both convex and non-convex clusters. Kd-tree based nearest neighbor search is used to reduce the complexity of finding the closest symmetric point. Adaptive mutation and crossover probabilities are used. The proposed GA with point symmetry (GAPS) distance based clustering algorithm is able to detect any type of clusters, irrespective of their geometrical shape and overlapping nature, as long as they possess the characteristic of symmetry. GAPS is compared with existing symmetry-based clustering technique SBKM, its modified version, and the well-known K-means algorithm. Sixteen data sets with widely varying characteristics are used to demonstrate its superiority. For real-life data sets, ANOVA and MANOVA statistical analyses are performed.