Search

The ongoing global pandemic of COVID-19, caused by SARS-CoV-2 has killed more than 5.9 million individuals out of ∼43 million confirmed infections. At present, several parts of the world are encountering the 3rd wave. Mass vaccination has been started in several countries but they are less likely to be broadly available for the current pandemic, repurposing of the existing drugs has drawn highest attention for an immediate solution. A recent publication has mapped the physical interactions of SARS-CoV-2 and human proteins by affinity-purification mass spectrometry (AP-MS) and identified 332 high-confidence SARS-CoV-2-human protein-protein interactions (PPIs). Here, we taken a network biology approach and constructed a human protein-protein interaction network (PPIN) with the above SARS-CoV-2 targeted proteins. We utilized a combination of essential network centrality measures and functional properties of the human proteins to identify the critical human targets of SARS-CoV-2. Four human proteins, namely PRKACA, RHOA, CDK5RAP2, and CEP250 have emerged as the best therapeutic targets, of which PRKACA and CEP250 were also found by another group as potential candidates for drug targets in COVID-19. We further found candidate drugs/compounds, such as guanosine triphosphate, remdesivir, adenosine monophosphate, MgATP, and H-89 dihydrochloride that bind the target human proteins. The urgency to prevent the spread of infection and the death of diseased individuals has prompted the search for agents from the pool of approved drugs to repurpose them for COVID-19. Our results indicate that host targeting therapy with the repurposed drugs may be a useful strategy for the treatment of SARS-CoV-2 infection.

In recent times, advanced data mining research has been mostly focusing on clustering of categorical data, where a natural ordering in attribute values is missing. To address this fact the Rough Fuzzy K-Modes clustering technique has been recently developed in order to handle imperfect information, i.e. indiscernibility (coarseness) and vagueness within the dataset. However, it has been observed that the said technique suffers from the problem of local optima due to the random choice of initial cluster modes. Hence, in this paper, we have proposed an integrated clustering technique using multi-phase learning. In this regard, first, Simulated Annealing based Rough Fuzzy K-Modes and Genetic Algorithm based Rough Fuzzy K-Modes are proposed in order to perform the clustering better by considering clustering as an underlying optimization problem. These clustering methods individually produce clusters having set of central and peripheral points. Thereafter, for each case, final improved clustering results are obtained by assigning peripheral points to a particular crisp cluster using Random Forest, where central points are used as training set. Second, the varying cardinality of the training and testing sets produced by each clustering method further motivated us to propose a generalized technique called Integrated Rough Fuzzy Clustering using Random Forest, where, results of three aforementioned clustering techniques are used to compute the roughness measure. Based on this measure, three different sets namely best central points, semi-best central points and pure peripheral points are determined. Thereafter, using multi-phase learning, best central points are used to classify the semi-best central points and then using both of them, pure peripheral points are classified by Random Forest. Experimental results are reported quantitatively and visually to demonstrate the effectiveness of the proposed methods in comparison with well-known state-of-the-art methods for six synthetic and five real-life datasets. Finally, statistical significance tests are conducted to establish the superiority of the results produced by the proposed methods.

Feature selection algorithm contributes a lot in the domain of medical diagnosis. Choosing a small subset of genes that enable a classifier to predict the presence or type of disease accurately is a difficult optimisation problem due to the size of the microarray data. The dual task of achieving higher accuracy and a small number of features makes it a challenging research problem. In our work, we have developed a Recursive Memetic Algorithm (RMA) model for selection of genes. It is a variant of Memetic Algorithm (MA) and performs much better than MA as well as Genetic Algorithm (GA). RMA has been applied on seven microarray datasets namely, AMLGSE2191, Colon, DLBCL, Leukaemia, Prostate, MLL and SRBCT. Encouraging results obtained by the proposed model, reported in this article, are biologically validated with the use of Gene Oncology, KEGG pathways and heat maps.

Biomedical image segmentation is essential for computerized medical image analysis. Deep learning algorithms allow us to design state-of-the-art models for solving segmentation problems. The U-Net and its variants have provided positive results across various datasets. However, the existing networks have the same receptive field at each level and the models are supervised only at the shallow level. Considering these two ideas, we have proposed the D3MSU-Net where the field of view in each level is varied depending upon the depth of the resolution layer and the model is supervised at each resolution level. We have evaluated our network in eight benchmark datasets such as Electron Microscopy, Lung segmentation, Montgomery Chest X-ray, Covid-Radiopaedia, Wound, Medetec, Brain MRI, and Covid-19 lung CT dataset. Additionally, we have provided the performance for various ablations. The experimental results show the superiority of the proposed network. The proposed D3MSU-Net and ablation models are available at www.github.com/shirshabose/D3MSUNET.

Personal vehicles are invariably being preferred over public transport nowadays. Contact-less feature inspection and analysis based on personal preferences will be in high demand among customers in the post-pandemic world. A comprehensive online car recommendation system will be the customers’ spontaneous choice to understand and select the features of vehicles. However, the clustering of such categorical features is a challenging task as it is difficult to compare two textual attributes. In this paper, we have designed a cloud-based system that will automatically address this issue. Motivated by the cooperative game theory and fuzzy technique, and integrating the concept of Shapley theorem, a categorical data clustering algorithm has been developed. At the same time, to overcome the major limitation of having a high time complexity of the order O ( n 2 ) associated with the Shapley computation, the proposed algorithm has been distributed using Apache Spark’s Map Reduce architecture in Google Cloud Platform. The model has been thoroughly validated based on its performance on several synthetic as well as real data sets. Finally, a car recommendation system has been proposed and tested on three car sell data sets. The proposed approach outperforms the corresponding existing categorical clustering approaches in terms of various clustering validity indices. To the best of the authors’ knowledge, this is the first attempt to apply Map Reduce based Shapley computation over the categorical clustering, which can find its application beyond the proposed car recommendation system as well.

Microarray technology has an important role in evaluating gene expression data with unique patterns into existence. In gene-expression based experiments, the expression level of the gene is constantly monitored in order to classify a tissue sample. In microarray technology, the expressions of the genes are altered with respect to pathogenes. The altered expression values can be identified by analyzing the genes of the tissue/cell that are affected along with the tissues/cells that are unaffected are termed as biomarkers. In the current paper, we have developed an Active Learning (AL) model by using Support Vector Machine (SVM) in association with feature-selection (FS) algorithm; called Symmetrical Uncertainty (SU) for the prediction of cancer. The effectiveness of the proposed AL and SU combination is manifested and the biomarkers or cancerous genes identified by the proposed method on four gene-expression data sets are reported. In addition, the biological significance tests are performed for the cancer biomarkers obtained from the data sets.

Graphical abstractDisplay Omitted Thresholding is a commonly used simple and effective technique for image segmentation. The computational time in multi-level thresholding significantly increases with the level of computation because of exhaustive searching, adding to exponential growth of computational complexity. Hence, in this paper, the features of quantum computing are exploited to introduce four different quantum inspired meta-heuristic techniques to accelerate the execution of multi-level thresholding. The proposed techniques are Quantum Inspired Genetic Algorithm, Quantum Inspired Simulated Annealing, Quantum Inspired Differential Evolution and Quantum Inspired Particle Swarm Optimization. The effectiveness of the proposed techniques is exhibited in comparison with the backtracking search optimization algorithm, the composite DE method, the classical genetic algorithm, the classical simulated annealing, the classical differential evolution and the classical particle swarm optimization for ten real life true colour images. The experimental results are presented in terms of optimal threshold values for each primary colour component, the fitness value and the computational time (in seconds) at different levels. Thereafter, the quality of thresholding is judged in terms of the peak signal-to-noise ratio for each technique. Moreover, statistical test, referred to as Friedman test, and also median based estimation among all techniques, are conducted separately to judge the preeminence of a technique among them. Finally, the performance of each technique is visually judged from convergence plots for all test images, which affirms that the proposed quantum inspired particle swarm optimization technique outperforms other techniques.

This paper proposes a novel method, Multi-Size patch based Collaborative Representation based Classification (CRC) strategy by Enhanced Ensemble Learning, for palm dorsa vein pattern (PDVP) based human recognition employing thermal imaging. This thermal PDVP imaging based human recognition methodology has been specifically employed to encounter the challenging crisis of intrusion posed by imposters. To address the Small Sample Size (SSS) problem, intrinsic to many biometric applications, each image is transformed into multiple patches, leading to an increase in the total number of samples. In a bid to make the classification strategy less sensitive to the choice of patch-size, the present paper proposes a new enhanced ensemble learning for the patch based CRC (PCRC) algorithm, where margin maximization is carried out using exponential squared loss minimization. This work also proposes how this loss minimization can be achieved by a stochastic optimization algorithm and solves this problem using artificial bee colony (ABC) algorithm. In this context, a new ABC variant, called modified interactive artificial bee colony (MI-ABC) algorithm, has also been proposed, which has been demonstrated to outperform the basic ABC and its several modern variants. The proposed methodology has been implemented on a wellstructured real database, developed in our laboratory using real subjects, and the results obtained in implementation phase clearly demonstrate that our proposed method could outperform its several competitors and achieve substantially high classification rates, for the problem under consideration.

Prostate cancer is the second most common type of cancer among men after skin cancer In this work, we present a comprehensive view on genomic and epigenomic changes following the incremental biological functionality. For gene selection, a new Feature Selection algorithm called Quantum Squirrel inspired Feature Selection is proposed here. While exploring the feature space, the proposed algorithm exploits the benefits of Squirrel Search Algorithm (a recently proposed swarm intelligence algorithm) along with Quantum mechanics. Moreover, a modified version of the end of winter concept is used to achieve effective dimension reduction capacity. Quantum Squirrel inspired Feature Selection is executed on both expression and methylation data of prostate cancer. The major challenge in gene selection is to bring down the number of selected features without compromising on accuracy. The proposed algorithm consistently achieves this goal and outperforms other state-of-the-art algorithms. The proposed algorithm has steadily attained 100% accuracy while selecting a much lower number of features (around 4), which is a major improvement over others. The top selected genes are biologically validated in terms of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and Gene Ontologies (GO), which further demonstrates the usefulness of the proposed method. The genes selected by Quantum Squirrel inspired Feature Selection show an association with prostate carcinoma and most are known biomarkers. A few novel biomarkers selected by proposed algorithm have also been detailed in this work. Source code of this work is available at: Quantum Squirrel inspired Feature Selection .

Breast cancer is the second leading cancer type among females. In this regard, it is found that microRNAs play an important role by regulating the gene expressions at the post-transcriptional phase. However, identification of the most influencing miRNAs in breast cancer subtypes is a challenging task, while the recent advancement in Next Generation Sequencing techniques allows analyzing high throughput expression data of miRNAs. Thus, we have conducted this research with the help of NGS data of breast cancer in order to identify the most significant miRNA biomarkers. The selected miRNA biomarkers are highly associated with the multiple breast cancer subtypes. For this purpose, a two-phase technique, called Machine Learning Integrated Ensemble of Feature Selection Methods, followed by survival analysis, is proposed. In the first phase, we have selected the best among seven machine learning techniques based on classification accuracy using the entire set of features (in this case miRNAs). Subsequently, eight different feature selection methods are used separately in order to rank the features and validate each set of top features using the selected machine learning technique by considering a multi-class classification task of the breast cancer subtypes. In the second phase, based on the classification accuracy values, the top features from each feature selection method are considered to make an ensemble to provide further categorization of the miRNAs as 8*, 7* up to 1*. The 8* miRNAs provide the highest average classification accuracy of 86% after 10-fold cross-validation. Thereafter, 27 miRNAs are identified from the list that is confined within 8* to 4* miRNAs based on their importance in survival for breast cancer subtypes using Cox regression based survival analysis. Moreover, expression analysis, regulatory network analysis, protein-protein interaction analysis, KEGG pathway and gene ontology enrichment analysis are performed in order to validate biological significance of the proposed solution. Additionally, we have prepared a miRNA-protein-drug interaction network to identify possible drug for the selected miRNAs. Thus, our findings may be considered during a clinical trial for the treatment of breast cancer patients.

Graphical abstractDisplay Omitted Pixel clustering in spectral domain is an important approach for the soft-tissue categorization of magnetic resonance (MR) brain images. In this regard, clustering algorithms based on type-1 fuzzy set theory are suitable for the overlapping partitions while the rough set based clustering algorithms deal with uncertainty and vagueness. However, additional degree of fuzziness makes the clustering more challenging for various subtle uncertainties and noisy data in the overlapping areas. Hence, this fact motivates us to propose a hybrid technique, called Rough Possibilistic Type-2 Fuzzy C-Means clustering with the integration of Random Forest. In the proposed method, possibilistic approach handles the noisy data better, whereas the other various uncertainties and inherent vagueness are taken care by type-2 fuzzy set and rough set theories. After clustering, it produces rough and crisp points. Thereafter, such crisp points are used to train the Random Forest classifier in order to classify the rough points for yielding better clustering solution. The performance of the proposed method has been demonstrated in comparison with several other recently proposed methods for MR brain image segmentation. Finally, superiority of the results produced by the proposed hybrid method has also been validated through statistical significance test.

Graphical abstractDisplay Omitted The efficient meta-heuristic techniques, called ant colony optimization, differential evolution and particle swarm optimization, inspired by the fundamental features of quantum systems, are presented in this paper. The proposed techniques are Quantum Inspired Ant Colony Optimization, Quantum Inspired Differential Evolution and Quantum Inspired Particle Swarm Optimization for Multi-level Colour Image Thresholding. These techniques find optimal threshold values at different levels of thresholding for colour images. A minimum cross entropy based thresholding method, called Li's method is employed as an objective (fitness) function for this purpose. The efficiency of the proposed techniques is exhibited computationally and visually on ten real life true colour images. Experiments with the composite DE (CoDE) method, the backtracking search optimization algorithm (BSA), the classical ant colony optimization (ACO), the classical differential evolution (DE) and the classical particle swarm optimization (PSO), have also been conducted subsequently along with the proposed techniques. Experimental results are described in terms of the best threshold value, fitness measure and the computational time (in seconds) for each technique at various levels. Thereafter, the accuracy and stability of the proposed techniques are established by computing the mean and standard deviation of fitness values for each technique. Moreover, the quality of thresholding for each technique is determined by computing the peak signal-to-noise ratio (PSNR) values at different levels. Afterwards, the statistical superiority of the proposed techniques is proved by incorporating Friedman test (statistical test) among different techniques. Finally, convergence curves for different techniques are presented for all test images to show the visual representation of results, which proves that the proposed Quantum Inspired Ant Colony Optimization technique outperforms all the other techniques.

Kernel based learning algorithms are sensitive to the choice of appropriate kernel function and parameter setting. Classification accuracies yielded by the kernel based classifiers may show variation depending on the choice of the kernel and its associated parameters. Suggesting an efficient kernel function and effective setting of kernel parameters are thus important problems for kernel based classifiers. In this article, we have investigated the performance of the existing kernel functions with our proposed kernel using support vector machines (SVMs). Linear, polynomial, sigmoid, radial basis function (RBF) and the proposed kernel are applied for the classification of the ten real life data sets having features and classes ranging from 4 to 19 and 2 to 5 respectively. The performance of different kernels is also demonstrated on two multispectral and two hyperspectral images. Experimental results on these data sets show the effectiveness of the proposed kernel on several data distributions.

Involvement of intrinsically disordered proteins (IDPs) with various dreadful diseases like cancer is an interesting research topic. In order to gain novel insights into the regulation of IDPs, in this article, we perform a transcriptomic analysis of mRNAs (genes) for transcripts encoding IDPs on a human multi-omics prostate carcinoma dataset having both gene expression and methylation data. In this regard, firstly the genes that consist of both the expression and methylation data, and that are corresponding to the cancer-related prostate-tissue-specific disordered proteins of MobiDb database, are selected. We apply standard t-test for determining differentially expressed genes as well as differentially methylated genes. A network having these genes and their targeter miRNAs from Diana Tarbase v7.0 database and corresponding Transcription Factors from TRANSFAC and ITFP databases, is then built. Thereafter, we perform literature search, and KEGG pathway and Gene Ontology analyses using DAVID database. Finally, we report several significant potential gene-markers (with the corresponding IDPs) that have inverse relationship between differential expression and methylation patterns, and that are hub genes of the TF-miRNA-gene network.

Display Omitted We develop a biomolecule-ranking method using (statistical) eigenvector centrality.We apply the ranking method on a multi-informative uterine leiomyoma dataset.Our ranking method is better than other ranking methods for TF-miRNA-gene network.We provide some top potential gene-markers (PTGS2 and TACSTD2) for the uterine tumor. Gene ranking is an important problem in bioinformatics. Here, we propose a new framework for ranking biomolecules (viz., miRNAs, transcription-factors/TFs and genes) in a multi-informative uterine leiomyoma dataset having both gene expression and methylation data using (statistical) eigenvector centrality based approach. At first, genes that are both differentially expressed and methylated, are identified using Limma statistical test. A network, comprising these genes, corresponding TFs from TRANSFAC and ITFP databases, and targeter miRNAs from miRWalk database, is then built. The biomolecules are then ranked based on eigenvector centrality. Our proposed method provides better average accuracy in hub gene and non-hub gene classifications than other methods. Furthermore, pre-ranked Gene set enrichment analysis is applied on the pathway database as well as GO-term databases of Molecular Signatures Database with providing a pre-ranked gene-list based on different centrality values for comparing among the ranking methods. Finally, top novel potential gene-markers for the uterine leiomyoma are provided.

Dominance is a fundamental concept in Multiobjective Optimization (MOO) where a solution is said to dominate the other if it is better in at least one objective, and same as or better in the other objectives. Given a pair of solutions, if neither dominates the other, then the solutions are said to be non-dominated. In real spaces, a large number of non-dominated solutions can exist in principle, making convergence of MOO techniques rather slow. Relaxed forms of dominance have been proposed in the literature for faster convergence. One such form is the ? -dominance, which in principle, partitions the space into ? -sized grids. The value of ? is often user defined and the performance of ? -dominance based multiobjective evolutionary algorithms ( ? -MOEAs) critically depends on the value of ? . In this article we propose a novel way of determining the value of ? , which is different for each objective function, based on the correlation between them. We call this approach Priority Based ? (PBE) as the method ranks the objectives according to their priority before calculating the ? values. PBE is incorporated in an archived simulated annealing based MOO technique called AMOSA. AMOSA has been earlier shown to outperform several well-known MOO techniques especially for many objective optimization. PBE based AMOSA, referred to as PBE-AMOSA, is found to comprehensively outperform AMOSA, MOEA/D-DE, the conventional ? based version ? -AMOSA and ? -MOEA both in case of benchmark test problems and 0/1 knapsack problem.

Categorical data is different from continuous data, where the values of attribute do not follow any natural ordering. Moreover, inherent complexities like uncertainty, vagueness and overlapping among clusters make the analysis of real life categorical data set more difficult. Recent literature review shows that the well-known categorical data clustering techniques are using different similarity/dissimilarity measures to tackle the inherent complexities of the categorical attribute values. Generally, it is hard to find single method and cluster validity measure that can be used as perfect or standard for all kinds of categorical data sets. Hence, in this paper first, a clustering method for categorical data is proposed by fusing rough set and fuzzy set theories. Subsequently, an ensemble based framework is designed with the recently proposed similarity/dissimilarity measures in order to have better clustering results for different types of categorical data sets. For this purpose, the proposed rough fuzzy clustering method is used sequentially with the integration of different measures to evolve the clustering solutions. Using consensus of these solutions, pure classified, semi rough and pure rough points are identified. Thereafter, machine learning method, called Random Forest, is used in incremental way to classify the semi and pure rough points using pure classified points to yield better clustering results. The performance of the proposed method has been demonstrated in comparison with several other recently developed clustering methods. Additionally, the selection of Random Forest in the proposed framework is justified by comparing its performance with other well-known machine learning methods like K-Nearest Neighbor and Support Vector Machine. Ten categorical data sets are used for the experimental purpose. Finally, statistical significance test has been conducted to judge the superiority of the results.

In real life, availability of correctly labeled data and handling of categorical data are often acknowledged as two major challenges in pattern analysis. Thus, clustering techniques are employed on unlabeled data to group them according to homogeneity. However, clustering techniques fail to make a decision while data are uncertain, ambiguous, vague, coincidental and overlapping in nature. Hence, in this case, the use of semi supervised technique can be useful. On the other hand, real life datasets are majorly categorical in nature, where natural ordering in attribute values is missing. This special property of categorical values with the inherent characteristics like uncertainty, ambiguity and vagueness makes clustering more complicated than numerical data. In recent times, credibilistic measure shows better performance over fuzzy and possibilistic measures while considering similar inherent characteristics in numerical data. Thus, these facts motivated us to propose a semi supervised clustering technique using credibilistic measure with the integration of machine learning techniques to address the above mentioned challenges of clustering categorical data. This semi supervised technique first clusters the dataset into K subsets with the proposed Credibilistic K -Mode, where credibilistic measure helps to determine the homogeneity by avoiding coincident clustering problem as well as finds the points those are certain to the clusters. Thereafter, in the second part of the semi supervised technique, clustered dataset is used to build a supervised model for classification of other unlabeled or uncertain data. This technique not only handles the unlabeled data better, but also yields improved results for uncertain or ambiguous data e.g, if the credibilistic measure is same for a data point in multiple classes. The results of the proposed technique are demonstrated quantitatively and visually in comparison with widely used state-of-the-art methods for eight synthetic and four real life datasets. Finally, statistical tests have been conducted to judge the statistical significance of the results produced by the proposed technique.

Image thresholding is well accepted and one of the most imperative practices to accomplish image segmentation. This has been widely studied over the past few decades. However, as the multi-level thresholding computationally takes more time when level increases, hence, in this article, quantum mechanism is used to propose six different quantum inspired meta-heuristic methods for performing multi-level thresholding faster. The proposed methods are Quantum Inspired Genetic Algorithm, Quantum Inspired Particle Swarm Optimization, Quantum Inspired Differential Evolution, Quantum Inspired Ant Colony Optimization, Quantum Inspired Simulated Annealing and Quantum Inspired Tabu Search. As a sequel to the proposed methods, we have also conducted experiments with the two-Stage multithreshold Otsu method, maximum tsallis entropy thresholding, the modified bacterial foraging algorithm, the classical particle swarm optimization and the classical genetic algorithm. The effectiveness of the proposed methods is demonstrated on fifteen images at the different level of thresholds quantitatively and visually. Thereafter, the results of six quantum meta-heuristic methods are considered to create consensus results. Finally, statistical test, called Friedman test, is conducted to judge the superiority of a method among them. Quantum Inspired Particle Swarm Optimization is found to be superior among the proposed six quantum meta-heuristic methods and the other five methods are used for comparison. A Friedman test again conducted between the Quantum Inspired Particle Swarm Optimization and all the other methods to justify the statistical superiority. Finally, the computational complexities of the proposed methods have been elucidated for the sake of finding out the time efficiency of the proposed methods.

The problem of clustering categorical data, where attribute values cannot be naturally ordered as numerical values, has gained more importance in recent time. Due to the special properties of categorical attributes, the clustering of categorical data seems to be more complicated than that of numerical data. Although, a few clustering algorithms that optimize single clustering objective have been proposed. It has found that such single measure may not be appropriate for all kind of datasets. Hence, in this article, an Incremental Learning based Multiobjective Fuzzy Clustering for Categorical Data is proposed. For this purpose, a multiobjective modified differential evolution based fuzzy clustering algorithm is developed. Thereafter, it integrates with the well-known supervised classifier, called random forest, using incremental learning to propose the aforementioned technique. Here, the multiobjective algorithm produces a set of optimal clustering solutions, known as Pareto-optimal solutions, by optimizing two conflicting objectives simultaneously. Subsequently, through incremental learning using random forest classifier final solution is evolved from the ensemble Pareto-optimal solutions. The results of the proposed method are demonstrated quantitatively and visually in comparison with widely used state-of-the-art methods for six synthetic and four real life datasets. Finally, statistical test is conducted to show the superiority of the results produced by the proposed method.

In this article a dense subgraph finding approach is adopted for the unsupervised feature selection problem. The feature set of a data is mapped to a graph representation with individual features constituting the vertex set and inter-feature mutual information denoting the edge weights. Feature selection is performed in a two-phase approach where the densest subgraph is first obtained so that the features are maximally non-redundant among each other. Finally, in the second stage, feature clustering around the non-redundant features is performed to produce the reduced feature set. An approximation algorithm is used for the densest subgraph finding. Empirically, the proposed approach is found to be competitive with several state of art unsupervised feature selection algorithms.

In this paper, two meta-heuristics techniques have been employed to introduce two new quantum inspired meta-heuristic methods, namely quantum inspired genetic algorithm and quantum inspired particle swarm optimization for bi-level thresholding. The proposed methods use Otsu's method, Ramesh's method, Li's method, Shanbag's method and also correlation coefficient as evaluation functions to determine optimal threshold values of gray-level images. They exploit the trivial concepts of quantum computing such as qubits and superposition of states. These properties help to exhibit the feature of parallelism that in turn utilizes the time discreteness of quantum mechanical systems. The proposed methods have been compared with their classical counterparts and later with the quantum evolutionary algorithm (QEA) proposed by Han et al. to evaluate the performance among all participating algorithms for three test images. The optimal threshold value with the corresponding fitness value, standard deviation of fitness and finally the computational time of each method for each test image have been reported. The results prove that the proposed methods are time efficient while compared to their conventional counterparts. Another comparative study of the proposed methods with the quantum evolutionary algorithm (QEA) proposed by Han et al. also reveals the weaknesses of the latter.

Recently, a new model of multiobjective simulated annealing, AMOSA, was developed which was found to provide improved performance for several multi objective optimization problems especially for problems with many objectives. In this article, we aim to further improve the performance of AMOSA by incorporating the concept of @e-dominance which is a more generalized form of conventional dominance. This strategy is referred to as @e-AMOSA. The result of @e-AMOSA is compared with those of AMOSA, NSGA-II and @e-MOEA and AMOSA for several test problems with number of objectives varying from two to fifteen and the number of variables varying from one to thirty. The performance of @e-AMOSA is also compared with other strategies for multiobjective 0/1 knapsack problem. A real life application of @e-AMOSA for clustering genes from gene expression data is also demonstrated. The results demonstrate the effectiveness of @e-AMOSA.

Land cover classification using remotely sensed data requires robust classification methods for the accurate mapping of complex land cover area of different categories. In this regard, support vector machines (SVMs) have recently received increasing attention. However, small number of training samples remains a bottleneck to design suitable supervised classifiers. On the other hand, adequate number of unlabeled data is available in remote sensing images which can be employed as additional source of information about margins. To fully leverage all of the precious unlabeled data, integration of filtering in a transductive SVM is proposed. Using two labeled image datasets of small size and two large unlabeled image datasets, the effectiveness of the proposed method is explored. Experimental results show that the proposed technique achieves average overall accuracies of around 4.5–7.8%, 0.8–2.6% and 0.9–2.2% more than the standard inductive SVM (ISVM), progressive transductive SVM (PTSVM) and low density separation (LDS) classifiers, respectively on larger domains in case of labeled datasets. Using image datasets, visual interpretation from the classified images as well as the segmentation quality reveal that the proposed method can efficiently filter informative data from the unlabeled samples.

The parallel self-organizing neural network (PSONN) architecture uses bilevel sigmoidal activation functions for the purpose of extraction of embedded objects from pure color noisy perspectives. The process of extraction often involves an enhancement of the images under consideration. The network employs multilevel sigmoidal activation function to segment true color images. Both these activation functions are characterized by fixed thresholding parameters, which do not incorporate the underlying heterogeneity in the image intensity gamut. Methods for incorporating dynamic thresholding mechanisms into the thresholding characteristics of the PSONN architecture are investigated in this paper. We also propose a parallel bi-directional self-organizing neural network (PBDSONN) architecture to address the limitations of the PSONN architecture. Three constituent BDSONNs in the proposed architecture process color component information by embedded adaptive fuzzy context sensitive thresholding (CONSENT) mechanisms. A source layer feeds the BDSONNs with input color component information. Another sink layer fuses the processed color component information into resultant outputs. Comparative results of the quality of the extracted/segmented images indicate the efficacy of the proposed PBDSONN architecture over the PSONN architecture with fixed as well as dynamic thresholding mechanisms.

In recent year, the problem of clustering in microarray data has been gaining significant attention. However most of the clustering methods attempt to find the group of genes where the number of cluster is known a priori. This fact motivated us to develop a new real-coded improved differential evolution based automatic fuzzy clustering algorithm which automatically evolves the number of clusters as well as the proper partitioning of a gene expression data set. To improve the result further, the clustering method is integrated with a support vector machine, a well-known technique for supervised learning. A fraction of the gene expression data points selected from different clusters based on their proximity to the respective centers, is used for training the SVM. The clustering assignments of the remaining gene expression data points are thereafter determined using the trained classifier. The performance of the proposed clustering technique has been demonstrated on five gene expression data sets by comparing it with the differential evolution based automatic fuzzy clustering, variable length genetic algorithm based fuzzy clustering and well known Fuzzy C-Means algorithm. Statistical significance test has been carried out to establish the statistical superiority of the proposed clustering approach. Biological significance test has also been carried out using a web based gene annotation tool to show that the proposed method is able to produce biologically relevant clusters of genes. The processed data sets and the matlab version of the software are available at http://bio.icm.edu.pl/~darman/IDEAFC-SVM/.

In this article, the performance of three major components of soft computing viz., genetic algorithm, simulated annealing and differential evolution have been studied for developing fuzzy clustering of microarray gene expression data. Microarray technology permits simultaneous monitoring of the expression levels of a large number of genes over different time points. Clustering is an important unsupervised analysis tool in this domain for finding groups of co-expressed genes. In this regard, the performance of the well-known fuzzy c-means algorithm is also studied in addition to the above mentioned soft computing-based approaches. Subsequently, support vector machine, a well-known technique for supervised learning, is utilized to improve the result of the clustering techniques. For this purpose, a fraction of the data points selected from different clusters based on their proximity to the respective centers, is used for training the support vector machine. The cluster assignments of the remaining points are thereafter determined using the trained classifier. Two publicly available benchmark microarray data sets have been used for demonstrating the effectiveness of the proposed approaches. Biological significance tests have been performed to show that the proposed method is able to produce biologically relevant clusters of co-expressed genes.

This article presents a new multiobjective differential evolution based fuzzy clustering technique. Recent research has shown that clustering techniques that optimize a single objective may not provide satisfactory result because no single validity measure works well on different kinds of data sets. The fact motivated us to present a new multiobjective Differential Evolution based fuzzy clustering technique that encodes the cluster centres in its vectors and optimizes multiple validity measures simultaneously. In the final generation, it produces a set of non-dominated solutions, from which the user can relatively judge and pick up the most promising one according to the problem requirements. Superiority of the proposed method over its single objective versions, multiobjective version of classical differential evolution and genetic algorithm, well-known fuzzy C-means and average linkage clustering algorithms has been demonstrated quantitatively and visually for several synthetic and real life data sets. Statistical significance test has been conducted to establish the statistical superiority of the proposed multiobjective clustering approach. Finally, the proposed algorithm has been applied for segmentation of a remote sensing image to show its effectiveness in pixel classification.

In this article, we present a semisupervised support vector machine that uses self-training approach. We then construct an ensemble of semisupervised SVM classifiers to address the problem of pixel classification of remote sensing images. Semisupervised support vector machines (S^3VMs) are based on applying the margin maximization principle to both labeled and unlabeled samples. The ensemble of SVM classifiers recognizes the conceptual similarity between component classifiers from the same data source. The effectiveness of the proposed technique is first demonstrated for two numeric remote sensing data described in terms of feature vectors and then identifying different land cover regions in remote sensing imagery. Experimental results on these datasets show that employing this learning scheme can increase the accuracy level. The performance of the ensemble is compared with one of its component classifier and conventional SVM in terms of accuracy and quantitative cluster validity indices.

This article proposes a novel multiobjective real coded genetic fuzzy clustering scheme for segmentation of multispectral magnetic resonance image (MRI) of the human brain. The proposed technique is able to automatically evolve the number of clusters along with the clustering result. The multiobjective variable string length clustering technique encodes the cluster centers in its chromosomes and simultaneously optimizes the global fuzzy compactness and fuzzy separation among the clusters. In the final generation, it produces a set of non-dominated solutions, from which the best solution in terms of a recently proposed validity index I is chosen to be the final clustering solution. The corresponding chromosome length provides the number of clusters. The proposed method is applied on many simulated T1-weighted, T2-weighted and proton density-weighted normal and MS lesion MRI brain images. Superiority of the proposed method over K-means, Fuzzy C-means, Expectation Maximization, hierarchical clustering, Single Objective Genetic clustering and other recent multiobjective clustering algorithms has been demonstrated quantitatively. The automatic segmentation obtained by the proposed clustering technique is also compared with the available ground truth information.

Neural network based image segmentation techniques primarily focus on the selection of appropriate thresholding points in the image feature space. Research initiatives in this direction aim at addressing this problem of effective threshold selection for activation functions. Multilevel activation functions resort to fixed and uniform thresholding mechanisms. These functions assume homogeneity of the image information content. In this paper, we propose a collection of adaptive thresholding approaches to multilevel activation functions. The proposed thresholding mechanisms incorporate the image context information in the thresholding process. Applications of these mechanisms are demonstrated on the segmentation of real life multilevel intensity images using a self-supervised multilayer self-organizing neural network (MLSONN) and a supervised pyramidal neural network (PyraNet). We also present a bi-directional self-organizing neural network (BDSONN) architecture suitable for multilevel image segmentation. The architecture uses an embedded adaptive thresholding mechanism to a characteristic multilevel activation function. The segmentation efficiencies of the thresholding mechanisms evaluated using four unsupervised measures of merit, are reported for the three neural network architectures considered.

Microarray technology has made it possible to monitor the expression levels of many genes simultaneously across a number of experimental conditions. Fuzzy clustering is an important tool for analyzing microarray gene expression data. In this article, a real-coded Simulated Annealing (VSA) based fuzzy clustering method with variable length configuration is developed and combined with popular Artificial Neural Network (ANN) based classifier. The idea is to refine the clustering produced by VSA using ANN classifier to obtain improved clustering performance. The proposed technique is used to cluster three publicly available real life microarray data sets. The superior performance of the proposed technique has been demonstrated by comparing with some widely used existing clustering algorithms. Also statistical significance test has been conducted to establish the statistical significance of the superior performance of the proposed clustering algorithm. Finally biological relevance of the clustering solutions are established.

Recent advancement in microarray technology permits monitoring of the expression levels of a large set of genes across a number of time points simultaneously. For extracting knowledge from such huge volume of microarray gene expression data, computational analysis is required. Clustering is one of the important data mining tools for analyzing such microarray data to group similar genes into clusters. Researchers have proposed a number of clustering algorithms in this purpose. In this article, an attempt has been made in order to improve the performance of fuzzy clustering by combining it with support vector machine (SVM) classifier. A recently proposed real-coded variable string length genetic algorithm based clustering technique and an iterated version of fuzzy C-means clustering have been utilized in this purpose. The performance of the proposed clustering scheme has been compared with that of some well-known existing clustering algorithms and their SVM boosted versions for one simulated and six real life gene expression data sets. Statistical significance test based on analysis of variance (ANOVA) followed by posteriori Tukey-Kramer multiple comparison test has been conducted to establish the statistical significance of the superior performance of the proposed clustering scheme. Moreover biological significance of the clustering solutions have been established.

The problem of classifying an image into different homogeneous regions is viewed as the task of clustering the pixels in the intensity space. In particular, satellite images contain landcover types some of which cover significantly large areas, while some (e.g., bridges and roads) occupy relatively much smaller regions. Detecting regions or clusters of such widely varying sizes presents a challenging task. A modified differential evolution based fuzzy clustering technique, is proposed in this article. Real-coded encoding of the cluster centres is used for this purpose. Results demonstrating the effectiveness of the proposed technique are provided for several synthetic and real life data sets as well as for some benchmark functions. Different landcover regions in remote sensing imagery have also been classified using the proposed technique to establish its efficiency. Statistical significance tests have been performed to establish the superiority of the proposed algorithm.

A time efficient technique for real-time tracking of high-speed objects in a video sequence is presented in this article. The technique is primarily based on the segmentation of the optical flow field computed between the successive image frames of a video sequence, followed by the tracking of a detected point of interest (POI) within the segmented flow field. In the initial phase of the technique, the optical flow field between the first two successive image frames acquired from a video sequence, is computed. A fuzzy hostility index indicative of the degree of coherence of the moving objects in the image frames, is used to segment the optical flow field. This yields different coherent regions of interest (ROIs) in the segmented flow field. A POI is then detected in the different ROIs obtained. Tracking of the moving object is then carried out by computing the flow fields between predefined ROIs in the neighborhood of the detected POI in the subsequent image frames. Since the selected ROIs are smaller than the image frames, a fair amount of reduction in the time required for the computation of the optical flow field is achieved, thereby facilitating real-time operation. An application of the proposed technique is demonstrated on three video sequences of high-speed flying fighter aircrafts.

This paper describes a technique for clustering homogeneously distributed data in a peer-to-peer environment like sensor networks. The proposed technique is based on the principles of the K-Means algorithm. It works in a localized asynchronous manner by communicating with the neighboring nodes. The paper offers extensive theoretical analysis of the algorithm that bounds the error in the distributed clustering process compared to the centralized approach that requires downloading all the observed data to a single site. Experimental results show that, in contrast to the case when all the data is transmitted to a central location for application of the conventional clustering algorithm, the communication cost (an important consideration in sensor networks which are typically equipped with limited battery power) of the proposed approach is significantly smaller. At the same time, the accuracy of the obtained centroids is high and the number of samples which are incorrectly labeled is also small.

In this article, the effectiveness of variable string length genetic algorithm along with a recently developed fuzzy cluster validity index (PBMF) has been demonstrated for clustering a data set into an unknown number of clusters. The flexibility of a variable string length Genetic Algorithm (VGA) is utilized in conjunction with the fuzzy indices to determine the number of clusters present in a data set as well as a good fuzzy partition of the data for that number of clusters. A comparative study has been performed for different validity indices, namely, PBMF, XB, PE and PC. The results of the fuzzy VGA algorithm are compared with those obtained by the well known FCM algorithm which is applicable only when the number of clusters is fixed a priori. Moreover, another genetic clustering scheme, that also requires fixing the value of the number of clusters, is implemented. The effectiveness of the PBMF index as the optimization criterion along with a genetic fuzzy partitioning technique is demonstrated on a number of artificial and real data sets including a remote sensing image of the city of Kolkata.

In this article, a cluster validity index and its fuzzification is described, which can provide a measure of goodness of clustering on different partitions of a data set. The maximum value of this index, called the PBM-index, across the hierarchy provides the best partitioning. The index is defined as a product of three factors, maximization of which ensures the formation of a small number of compact clusters with large separation between at least two clusters. We have used both the k-means and the expectation maximization algorithms as underlying crisp clustering techniques. For fuzzy clustering, we have utilized the well-known fuzzy c-means algorithm. Results demonstrating the superiority of the PBM-index in appropriately determining the number of clusters, as compared to three other well-known measures, the Davies–Bouldin index, Dunn's index and the Xie–Beni index, are provided for several artificial and real-life data sets.

This paper presents a goal programming (GP) procedure for fuzzy multiobjective linear fractional programming (FMOLFP) problems. In the proposed approach, which is motivated by Mohamed (Fuzzy Sets and Systems 89 (1997) 215), GP model for achievement of the highest membership value of each of fuzzy goals defined for the fractional objectives is formulated. In the solution process, the method of variable change on the under- and over- deviational variables of the membership goals associated with the fuzzy goals of the model is introduced to solve the problem efficiently by using linear goal programming (LGP) methodology. The approach is illustrated by two numerical examples.

Nearest Neighbor rule is one of the most commonly used supervised classification procedures due to its inherent simplicity and intuitive appeal. However, it suffers from the major limitation of requiring n distance computations, where n is the size of the training data (or prototypes), for computing the nearest neighbor of a point. In this paper we suggest a simple approach based on rearrangement of the training data set in a certain order, such that the number of distance computations is significantly reduced. At the same time, the classification accuracy of the original rule remains unaffected. This method requires the storage of at most n distances in addition to the prototypes. The superiority of the proposed method in comparison to some other methods is clearly established in terms of the number of distances computed, the time required for finding the nearest neighbor, number of optimized operations required in the overhead computation and memory requirements. Variation of the performance of the proposed method with the size of the test data is also demonstrated.

A genetic algorithm-based efficient clustering technique that utilizes the principles of K-Means algorithm is described in this paper. The algorithm called KGA-clustering, while exploiting the searching capability of K-Means, avoids its major limitation of getting stuck at locally optimal values. Its superiority over the K-Means algorithm and another genetic algorithm-based clustering method, is extensively demonstrated for several artificial and real life data sets. A real life application of the KGA-clustering in classifying the pixels of a satellite image of a part of the city of Mumbai is provided.

In this article the searching capability of genetic algorithms has been exploited for automatically evolving the number of clusters as well as proper clustering of any data set. A new string representation, comprising both real numbers and the do not care symbol, is used in order to encode a variable number of clusters. The Davies–Bouldin index is used as a measure of the validity of the clusters. Effectiveness of the genetic clustering scheme is demonstrated for both artificial and real-life data sets. Utility of the genetic clustering technique is also demonstrated for a satellite image of a part of the city Calcutta. The proposed technique is able to distinguish some characteristic landcover types in the image.

An efficient scheme for fault tolerant mapping of permutations is designed. The proposed algorithm uses extra passes through the network, instead of additional hardware.