Framework for efficient feature selection in genetic algorithm based data mining

We present the design of more effective and efficient genetic algorithm based data mining techniques that use the concepts of feature selection. Explicit feature selection is traditionally done as a wrapper approach where every candidate feature subset is evaluated by executing the data mining algorithm on that subset. In this article we present a GA for doing both the tasks of mining and feature selection simultaneously by evolving a binary code along side the chromosome structure used for evolving the rules. We then present a wrapper approach to feature selection based on Hausdorff distance measure. Results from applying the above techniques to a real world data mining problem show that combining both the feature selection methods provides the best performance in terms of prediction accuracy and computational efficiency.     

A Hybrid Micro­Macro­Model for the Description of Evolving Anisotropy in Finite Polycrystal Plasticity

The paper outlines recent developments of an efficient computational micro-macro modeling of evolving anisotropies in metallic polycrystals. Main focus is put onto large strain deformation processes where the anisotropy is caused by developments of crystallographic texture. We construct a hybrid micro-macro framework that mixes ingredients of a purely macroscopic modeling with scale bridging operations of selected micromechanisms. On the micromechanical side, we develop a new algorithmic procedure to capture the crystal reorientation for evolving fcc and bcc textures based on a parametrization of rotations in the Rodigues space. The computational model provides a fast and robust method for the estimation of evolving textures. This computational tool for texture estimation is incorporated in a modular format into a micro-macro-model, where it governs the evolution of macrostructural tensors due to texture development. The general framework for the hybrid embedding is a purely phenomenological setting of anisotropic finite plasticity in the logarithmic strain space. The model provides an efficient and computationally handable two-scale approach for the prediction of effects caused by complex microstructural changes in polycrystals. The capability of the proposed method is demonstrated by means of representative numerical examples. (© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

Sparse <Emphasis Type="Italic">p</Emphasis>-adic data coding for computationally efficient and effective big data analytics

We develop the theory and practical implementation of p-adic sparse coding of data. Rather than the standard, sparsifying criterion that uses the L ₀ pseudo-norm, we use the p-adic norm.We require that the hierarchy or tree be node-ranked, as is standard practice in agglomerative and other hierarchical clustering, but not necessarily with decision trees. In order to structure the data, all computational processing operations are direct reading of the data, or are bounded by a constant number of direct readings of the data, implying linear computational time. Through p-adic sparse data coding, efficient storage results, and for bounded p-adic norm stored data, search and retrieval are constant time operations. Examples show the effectiveness of this new approach to content-driven encoding and displaying of data.     

Markov chain-based degree distributions of evolving networks

In this paper, we study a class of stochastic processes, called evolving network Markov chains, in evolving networks. Our approach is to transform the degree distribution problem of an evolving network to a corresponding problem of evolving network Markov chains. We investigate the evolving network Markov chains, thereby obtaining some exact formulas as well as a precise criterion for determining whether the steady degree distribution of the evolving network is a power-law or not. With this new method, we finally obtain a rigorous, exact and unified solution of the steady degree distribution of the evolving network.     

Scalar Correction Method for Solving Large Scale Unconstrained Minimization Problems

We introduce a gradient descent algorithm for solving large scale unconstrained nonlinear optimization problems. The computation of the initial trial steplength is based on the usage of both the quasi-Newton property and the Hessian inverse approximation by an appropriate scalar matrix. The nonmonotone line search technique for the steplength calculation is applied later. The computational and storage complexity of the new method is equal to the computational and storage complexity of the Barzilai and Borwein method. On the other hand, the reported numerical results indicate improvements in favor of the new method with respect to the well known global Barzilai and Borwein method.     

Solving the Simple Plant Location Problem using a Data Correcting Approach

The Data Correcting Algorithm is a branch and bound type algorithm in which the data of a given problem instance is `corrected' at each branching in such a way that the new instance will be as close as possible to a polynomially solvable instance and the result satisfies an acceptable accuracy (the difference between optimal and current solution). In this paper the data correcting algorithm is applied to determining exact and approximate optimal solutions to the simple plant location problem. Implementations of the algorithm are based on a pseudo-Boolean representation of the goal function of this problem, and a new reduction rule. We study the efficiency of the data correcting approach using two different bounds, the Khachaturov-Minoux bound and the Erlenkotter bound. We present computational results on several benchmark instances, which confirm the efficiency of the data-correcting approach.     

On a Class of Shrinkage Priors for Covariance Matrix Estimation

We propose a flexible class of models based on scale mixture of uniform distributions to construct shrinkage priors for covariance matrix estimation. This new class of priors enjoys a number of advantages over the traditional scale mixture of normal priors, including its simplicity and flexibility in characterizing the prior density. We also exhibit a simple, easy to implement Gibbs sampler for posterior simulation, which leads to efficient estimation in high-dimensional problems. We first discuss the theory and computational details of this new approach and then extend the basic model to a new class of multivariate conditional autoregressive models for analyzing multivariate areal data. The proposed spatial model flexibly characterizes both the spatial and the outcome correlation structures at an appealing computational cost. Examples consisting of both synthetic and real-world data show the utility of this new framework in terms of robust estimation as well as improved predictive performance. Supplementary materials are available online.     

Network reoptimization procedures for multiobjective network probelems

Network models are attractive because of their computational efficiency. Network applications can involve multiple objective analysis. Multiple objective analysis requires generating nondominated solutions in various forms. Two general methods exist to generate new solutions in continuous optimization: changing objective function weights and inserting objective bounds through constraints. In network flow problems, modifying weights is straightforward, allowing use of efficient network codes. Use of bounds on objective attainment levels can provide a more controlled generation of solutions reflecting tradeoffs among objectives. To constrain objective attainment, however, would require a side constrained network code, sacrificing some computational efficiency for greater model flexibility. We develop reoptimization procedures for the side constrained problem and use them in conjunction with simplex-based techniques. Our approach provides a useful tool for generating solutions allowing greater decision maker control over objective attainments, allowing multiobjective analysis of large-scale problems. Results are compared with solutions obtained from the computationally more attractive weighting technique. Reoptimization procedures are discussed as a means of more efficiently conducting multiple objective network analyses.     

The variable neighborhood search for the two machine flow shop problem with a passive prefetch

We consider the two machine flow shop scheduling problem with passive loading of the buffer on the second machine. To compute lower bounds for the global optimum, we present four integer linear programming formulations of the problem. Three local search methods with variable neighborhoods are developed for obtaining upper bounds. Some new large neighborhood is designed. Our methods use this neighborhood along with some other well-known neighborhoods. For computational experiments, we present a new class of test instances with known global optima. Computational results indicate a high efficiency of the proposed approach for the new class of instances as well as for other classes of instances of the problem.     

A collaboratory for radiation therapy treatment planning optimization research

Intensity modulated radiation therapy treatment planning (IMRTP) is a challenging application of optimization technology. We present software tools to facilitate IMRTP research by computational scientists who may not have convenient access to radiotherapy treatment planning systems. The tools, developed within Matlab and CERR (computational environment for radiotherapy research), allow convenient access, visualization, programmable manipulation, and sharing of patient treatment planning data, as well as convenient generation of dosimetric data needed as input for treatment plan optimization research. CERR/Matlab also provides a common framework for storing, reviewing, sharing, and comparing optimized dose distributions from multiple researchers.     

Geometry + Simulation Modules: Implementing Isogeometric Analysis

Isogeometric analysis (IGA) is a recently developed simulation method that allows integration of finite element analysis (FEA) with conventional computer-aided design (CAD) software [1,3]. This goal requires new software design strategies, in order to enable the use of CAD data in the analysis pipeline. To this end, we have initiated G + SMO (Geometry+Simulation Modules), an open-source, C++ library for IGA. G + SMO is an object-oriented, template library, that implements a generic concept for IGA, based on abstract classes for discretization basis, geometry map, assembler, solver and so on. It makes use of object polymorphism and inheritance techniques to provide a common framework for IGA, for a variety of different basis-types which are available. A highlight of our design is the dimension independent code, realized by means of template meta-programming. Some of the features already available include computing with B-spline, Bernstein, NURBS bases, as well as hierarchical and truncated hierarchical bases of arbitrary polynomial order. These basis functions are used in continuous and discontinuous Galerkin approximation of PDEs over (non-)conforming multi-patch computational (physical) domains. (© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

The single-assignment hub covering problem: Models and linearizations

We study the hub covering problem which, so far, has remained one of the unstudied hub location problems in the literature. We give a combinatorial and a new integer programming formulation of the hub covering problem that is different from earlier integer programming formulations. Both new and old formulations are nonlinear binary integer programs. We give three linearizations for the old model and one linearization for the new one and test their computational performances based on 80 instances of the CAB data set. Computational results indicate that the linear version of the new model performs significantly better than the most successful linearization of the old model both in terms of average and maximum CPU times as well as in core storage requirements.     

A Benders decomposition approach for a distribution network design problem with consolidation and capacity considerations

We develop a model for a strategic freight-forwarding network design problem in which the design decisions involve the locations and capacities of consolidation and deconsolidation centers, and capacities on linehaul linkages as well as the shipment routes from origins to destinations through centers. We devise a solution approach based on Benders decomposition and conduct a computational study that illustrates the efficiency and the effectiveness of the approach.     

Lower Dimensional Representation of Text Data Based on Centroids and Least Squares

Dimension reduction in today's vector space based information retrieval system is essential for improving computational efficiency in handling massive amounts of data. A mathematical framework for lower dimensional representation of text data in vector space based information retrieval is proposed using minimization and a matrix rank reduction formula. We illustrate how the commonly used Latent Semantic Indexing based on the Singular Value Decomposition (LSI/SVD) can be derived as a method for dimension reduction from our mathematical framework. Then two new methods for dimension reduction based on the centroids of data clusters are proposed and shown to be more efficient and effective than LSI/SVD when we have a priori information on the cluster structure of the data. Several advantages of the new methods in terms of computational efficiency and data representation in the reduced space, as well as their mathematical properties are discussed.Experimental results are presented to illustrate the effectiveness of our methods on certain classification problems in a reduced dimensional space. The results indicate that for a successful lower dimensional representation of the data, it is important to incorporate a priori knowledge in the dimension reduction algorithms.     

DCA based algorithms for multiple sequence alignment (MSA)

In the last years many techniques in bioinformatics have been developed for the central and complex problem of optimally aligning biological sequences. In this paper we propose a new optimization approach based on DC (Difference of Convex functions) programming and DC Algorithm (DCA) for the multiple sequence alignment in its equivalent binary linear program, called “Maximum Weight Trace” problem. This problem is beforehand recast as a polyhedral DC program with the help of exact penalty techniques in DC programming. Our customized DCA, requiring solution of a few linear programs, is original because it converges after finitely many iterations to a binary solution while it works in a continuous domain. To scale-up large-scale (MSA), a constraint generation technique is introduced in DCA. Preliminary computational experiments on benchmark data show the efficiency of the proposed algorithm DCAMSA, which generally outperforms some standard algorithms.     

Systems biology reveals biology of systems

In the last decades, genomic and postgenomic technologies obtained a great amount of information on molecular bases of cell physiology and organization. In spite of this, the knowledge of cells and living organisms in their entirety, is far from being achieved. In order to deal with biological complexity, Systems Biology uses a new approach to overcome this inadequacy. Despite different definitions, Systems Biology's view of biological phenomena highlights that a holistic perspective is needed to integrate and understand the huge amount of empirical data which have been collected. This is one of the aspects that makes Systems Biology so interesting, from a theoretical and epistemological point of view, and that renders it a useful tool to help students approach living beings' dynamics within a comprehensive framework of their biological features as well. © 2010 Wiley Periodicals, Inc. Complexity, 2010     

New higher-order numerical one-step methods based on the Adomian and the modified decomposition methods

We develop new, higher-order numerical one-step methods and apply them to several examples to investigate approximate discrete solutions of nonlinear differential equations. These new algorithms are derived from the Adomian decomposition method (ADM) and the Rach-Adomian-Meyers modified decomposition method (MDM) to present an alternative to such classic schemes as the explicit Runge-Kutta methods for engineering models, which require high accuracy with low computational costs for repetitive simulations in contrast to a one-size-fits-all philosophy. This new approach incorporates the notion of analytic continuation, which extends the region of convergence without resort to other techniques that are also used to accelerate the rate of convergence such as the diagonal Padé approximants or the iterated Shanks transforms. Hence global solutions instead of only local solutions are directly realized albeit in a discretized representation. We observe that one of the difficulties in applying explicit Runge-Kutta one-step methods is that there is no general procedure to generate higher-order numeric methods. It becomes a time-consuming, tedious endeavor to generate higher-order explicit Runge-Kutta formulas, because it is constrained by the traditional Picard formalism as used to represent nonlinear differential equations. The ADM and the MDM rely instead upon Adomian’s representation and the Adomian polynomials to permit a straightforward universal procedure to generate higher-order numeric methods at will such as a 12th-order or 24th-order one-step method, if need be. Another key advantage is that we can easily estimate the maximum step-size prior to computing data sets representing the discretized solution, because we can approximate the radius of convergence from the solution approximants unlike the Runge-Kutta approach with its intrinsic linearization between computed data points. We propose new variable step-size, variable order and variable step-size, variable order algorithms for automatic step-size control to increase the computational efficiency and reduce the computational costs even further for critical engineering models.     

A mathematical model of mast cell response to acupuncture needling

We introduce a new model of mast cell response to acupuncture needling based on the Keller–Segel model for chemotaxis. The needle manipulation induces the release of a chemoattractant by the mast cells. We show, in a simplified case, that blow-up of the solution occurs in finite time for large initial data concentrated around the acupoint. In those conditions, blow-up is the result of aggregation of cells and could indicate the efficiency of the acupuncture manipulation of the needle at one acupoint.     

<Emphasis Type="Italic">Graph_sampler</Emphasis>: a simple tool for fully Bayesian analyses of DAG-models

Bayesian networks (BNs) are widely used graphical models usable to draw statistical inference about directed acyclic graphs. We presented here Graph_sampler a fast free C language software for structural inference on BNs. Graph_sampler uses a fully Bayesian approach in which the marginal likelihood of the data and prior information about the network structure are considered. This new software can handle both the continuous as well as discrete data and based on the data type two different models are formulated. The software also provides a wide variety of structure prior which can depict either the global or local properties of the graph structure. Now based on the type of structure prior selected, we considered a wide range of possible values for the prior making it either informative or uninformative. We proposed a new and much faster jumping kernel strategy in the Metropolis–Hastings algorithm. The source C code distributed is very compact, fast, uses low memory and disk storage. We performed out several analyses based on different simulated data sets and synthetic as well as real networks to discuss the performance of Graph_sampler.