Search

The concept of opposition-based learning ( OBL ) was first introduced as a scheme for machine intelligence. In a very short period of time, some other variants of opposite numbers were proposed and opposition was applied to various research areas. In metaheuristic optimization algorithms, the main idea behind applying opposite numbers is the simultaneous consideration of a candidate solution and its corresponding opposite candidate in order to achieve a better approximation for the current solution. This paper proposes an opposition-based metaheuristic optimization algorithm ( OBA ) and a new and efficient opposition named comprehensive opposition ( CO ) as its main operator. In this paper it is mathematically proven that CO not only increases the chance of achieving better approximations for the solution but also guarantees the global convergence of OBA . The efficiency of the proposed method has been compared with some well-known heuristic search methods. The obtained results confirm the high performance of the proposed method in solving various function optimizations. [ABSTRACT FROM AUTHOR]

The detection and location of objects concealed under clothing is a very challenging task that has crucial applications in security. In this domain, passive millimeter-wave images (PMMWIs) can be used. However, the quality of the acquired images, and the unknown position, shape, and size of hidden objects render this task difficult. In this paper, we propose a machine learning-based solution to this detection/localization problem. Our method outperforms currently used approaches. The effect of non-stationary noise on different classification algorithms is analyzed and discussed, and a detailed experimental comparative study of classification techniques is presented using a new and comprehensive PMMWI database. The low computational testing cost of this solution allows for its use in real-time applications. [ABSTRACT FROM AUTHOR]

This paper presents a new general methodological approach to imbalanced learning as one of the challenging problems in pattern classification. The presented method is founded on maximization of the sample entropy. The method involves detection of distributive properties of ideally balanced regular lattice sample and acceptable transfer of these properties to an arbitrary imbalanced sample increasing its representativeness. The proposed procedure assumes undersampling applied on areas of high probability density in the sample space combined with oversampling in the areas of low density. The main achievement of this method is the increased sample class entropy which reduces the inductive learner’s tendency to favor prominent class, or cluster. In addition to class balancing, this method can be useful for function approximation, clustering, and sample dimension reduction. The high degree of generality of the method implies its applicability on data of various complexity and imbalance. The presented theoretical foundation of the method was verified on a set of proper synthetic samples. The method’s practical usability is confirmed by a comparative classification of a large set of databases including speech signal samples. [ABSTRACT FROM AUTHOR]

In this paper, a new sequential learning algorithm is constructed by combining the Online Sequential Extreme Learning Machine (OS-ELM) and Kalman filter regression. The Kalman Online Sequential Extreme Learning Machine (KOSELM) handles the problem of multicollinearity of the OS-ELM, which can generate poor predictions and unstable models. The KOSELM learns the training data one-by-one or chunk-by-chunk by adjusting the variance of the output weights through the Kalman filter. The performance of the proposed algorithm has been validated on benchmark regression datasets, and the results show that KOSELM can achieve a higher learning accuracy than OS-ELM and its related extensions. A statistical validation for the differences of the accuracy for all algorithms is performed, and the results confirm that KOSELM has better stability than ReOS-ELM, TOSELM and LS-IELM. [ABSTRACT FROM AUTHOR]

For years, meta-heuristic algorithms have been widely studied and many improved versions have been developed: from the evolution of the swarm topologies of the Particle Swarm Optimization algorithm, to the using of machine learning to Differential Evolutionary algorithms. However, the tuning of the fundamental meta-heuristic parameters has been less studied, but may lead to significant improvements on the convergence accuracy of these algorithms. This paper aims at developing an automated methodology to calibrate the parameters of population-based meta-heuristic algorithms for optimization problems. Based on the kriging estimation of the best combination of parameters, the Automated parameter tuning of Meta-heuristics (AptM) methodology gives the optimal algorithm setup for each considered problem in order to lead to a better convergence accuracy. The proposed AptM methodology is used to tune three different meta-heuristic algorithms, each applied to twelve mathematical unimodal or multimodal objective functions. AptM methodology performance is assessed by comparison of classical setups usually used in the literature. The numerical results show that the AptM methodology allows a significant improvement of the convergence accuracy of meta-heuristics with an average improvement of 62.02%, 69.12% and 64.94% on optimization problems defined in dimensions 10, 30 and 50 respectively. An experimental criterion is defined based on the convergence accuracy of the AptM methodology over the classical setups, assessing the AptM performances. The previous experimental criterion allows to compare the AptM methodology over the base-set. The AptM methodology shows a significant improvement of the algorithms performance on 97.2% of the tested problems. [ABSTRACT FROM AUTHOR]

In this paper, a multi-class classification method based on heterogeneous ensemble of one-class classifiers is proposed. The proposed method consists of two phases: training heterogeneous one-class classifiers for each class using various one-class classification algorithms, and constructing an ensemble by combining the base classifiers using multi-response linear regression-based stacking. The use of various classification algorithms contributes towards increasing the diversity of the ensemble, while stacking resolves the normalization issues on different scales of outputs obtained from the base classifiers. In addition, we also demonstrate the selective utilization of base classifiers by adopting a stepwise variable selection procedure during stacking. Through our experiments on multi-class benchmark datasets, we concluded that our proposed method outperforms the methods that are based on single one-class classification algorithms with statistical significance. [ABSTRACT FROM AUTHOR]

The paper presents a novel teaching–learning-based optimization (TLBO) algorithm, the Gaussian bare-bones TLBO (GBTLBO) algorithm, with its modified version (MGBTLBO) for the optimal reactive power dispatch (ORPD) problem with discrete and continuous control variables in the standard IEEE power systems for reduction in power transmission loss. The feasibility and performance of the GBTLBO and MGBTLBO algorithms are demonstrated for standard IEEE 14-bus and standard IEEE 30-bus systems. A comparison of simulation results reveals optimization efficacy of the GBTLBO and MGBTLBO algorithms over other well established other algorithms like bare-bones differential evolution (BBDE) and bare-bones particle swarm optimization (BBPSO) algorithm. Results for ORPD problem demonstrate superiority in terms of solution quality of the GBTLBO and MGBTLBO algorithms over original TLBO algorithm and other algorithm. [ABSTRACT FROM AUTHOR]

Gravitational Search Algorithm (GSA) is a stochastic population-based metaheuristic designed for solving continuous optimization problems. It has a flexible and well-balanced mechanism for enhancing exploration and exploitation abilities. In this paper, we adapt the structure of GSA for solving the data clustering problem, the problem of grouping data into clusters such that the data in each cluster share a high degree of similarity while being very dissimilar to data from other clusters. The proposed algorithm, which is called Grouping GSA (GGSA), differs from the standard GSA in two important aspects. First, a special encoding scheme, called grouping encoding, is used in order to make the relevant structures of clustering problems become parts of solutions. Second, given the encoding, special GSA updating equations suitable for the solutions with grouping encoding are used. The performance of the proposed algorithm is evaluated through several benchmark datasets from the well-known UCI Machine Learning Repository. Its performance is compared with the standard GSA, the Artificial Bee Colony (ABC), the Particle Swarm Optimization (PSO), the Firefly Algorithm (FA), and nine other well-known classical classification techniques from the literature. The simulation results indicate that GGSA can effectively be used for multivariate data clustering. [ABSTRACT FROM AUTHOR]

Abstract: Traditional cross-domain learning methods transfer learning from a source domain to a target domain. In this paper, we propose the multiple-domain learning problem for several equally treated domains. The multiple-domain learning problem assumes that samples from different domains have different distributions, but share the same feature and class label spaces. Each domain could be a target domain, while also be a source domain for other domains. A novel multiple-domain representation method is proposed for the multiple-domain learning problem. This method is based on nonnegative matrix factorization (NMF), and tries to learn a basis matrix and coding vectors for samples, so that the domain distribution mismatch among different domains will be reduced under an extended variation of the maximum mean discrepancy (MMD) criterion. The novel algorithm — multiple-domain NMF (MDNMF) — was evaluated on two challenging multiple-domain learning problems — multiple user spam email detection and multiple-domain glioma diagnosis. The effectiveness of the proposed algorithm is experimentally verified. [Copyright &y& Elsevier]

Abstract: Since given classification data often contains redundant, useless or misleading features, feature selection is an important pre-processing step for solving classification problems. This problem is often solved by applying evolutionary algorithms to decrease the dimensional number of features involved. Removing irrelevant features in the feature space and identifying relevant features correctly is the primary objective, which can increase classification accuracy. In this paper, a novel QBGSA–K-NN hybrid system which hybridizes the quantum-inspired binary gravitational search algorithm (QBGSA) with the K-nearest neighbor (K-NN) method with leave-one-out cross-validation (LOOCV) is proposed. The main aim of this system is to improve classification accuracy with an appropriate feature subset in binary problems. We evaluate the proposed hybrid system on several UCI machine learning benchmark examples. The experimental results show that the proposed method is able to select the discriminating input features correctly and achieve high classification accuracy which is comparable to or better than well-known similar classifier systems. [Copyright &y& Elsevier]

Deep learning-based target detection techniques have already made a wide-range impact on our daily life. Currently, a feature pyramid is a widely utilized technique for multiscale target detection, the effectiveness of the technique has already been proved. Nevertheless, in the pyramid structure, problems, such as multiscale feature alignment, model turmoil after fusion, feature redundancy, and no-local feature fusion, exist. In this paper, we propose a novel progressive structure network to solve the aforementioned problems. The proposed structure contains three modules: multiscale feature alignment fusion, different scale channels & spatial location adaptive weighted fusion, and multiscale global and local feature fusion. The proposed structure is capable of fusing information from different feature layers more effectively. Subsequently, the semantic gaps among different scales can be reduced. Furthermore, the proposed structure can maintain the stability of the detection network and its performance has been proved by comparing with other state-of-art feature fusion method. The proposed progressive network structure has also been applied to actual target detection tasks and the practical application effectiveness of our method has been verified. [ABSTRACT FROM AUTHOR]

The problem of identifying and therefore modelling a complex system makes use of various techniques and strategies whose computational efforts change drastically. It is not straightforward to analyse the complexity of a system as a whole because of myriads of factors, such as the way of arranging its constituent items and how they interact mutually. Intuitively, the bigger the set of sub-parts is, the more numerous the degrees of freedom are. Additionally there is not a specific and global criterion for optimally determining an always-working method that makes the identification procedure easier, especially in those contexts where the number of unknown variables can make the difference. In this sense, plasma physics is not an exception, being a field where complex phenomena, such as plasma instabilities, easily arise. From a systemic, high-level perspective, the possibility of employing a model that can describe these behaviours is particularly appealing, since it can be exploited for control applications that have not to neglect the underlying physical nature. So far, most of the work published in literature has focused on more physically-grounded models, which could describe how plasma physics works in detail, but very little has been done as mentioned before, with the aim of providing a computational, yet system-oriented, insight of these physical systems. Starting from real flux measurements recorded thanks to suitable sensors installed inside Tokamak machines, the paper attempts to provide a solution based on already known tools available in literature to solve the aforementioned problem, by combining both machine learning-based strategies for dimensionality reduction and control theory. More in detail, the whole architecture presented in this work is founded on the use of auto-encoders, which are intrinsically capable of compressing input features thanks to their structure, and Hammerstein–Wiener models, which are structurally endowed with both linear and non-linear sub-modellers for better capturing the whole dynamics to identify. By merging these functional blocks, it is possible to address both the issue of establishing the most relevant sub-set of variables for identification and the identification problem itself, resulting in a fully customisable approach to data-driven modelling. [ABSTRACT FROM AUTHOR]