Exponentiated gradient algorithms for conditional random fields and max-margin Markov Networks

Log-linear and maximum-margin models are two commonly-used methods in supervised machine learning, and are frequently used in structured prediction problems. Efficient learning of parameters in these models is therefore an important problem, and becomes a key factor when learning from very large data sets. This paper describes exponentiated gradient (EG) algorithms for training such models, where EG updates are applied to the convex dual of either the log-linear or max-margin objective function; the dual in both the log-linear and max-margin cases corresponds to minimizing a convex function with simplex constraints. We study both batch and online variants of the algorithm, and provide rates of convergence for both cases. In the max-margin case, O(1/ε) EG updates are required to reach a given accuracy ε in the dual; in contrast, for log-linear models only O(log(1/ε)) updates are required. For both the max-margin and log-linear cases, our bounds suggest that the online EG algorithm requires a factor of n less computation to reach a desired accuracy than the batch EG algorithm, where n is the number of training examples. Our experiments confirm that the online algorithms are much faster than the batch algorithms in practice. We describe how the EG updates factor in a convenient way for structured prediction problems, allowing the algorithms to be efficiently applied to problems such as sequence learning or natural language parsing. We perform extensive evaluation of the algorithms, comparing them to L-BFGS and stochastic gradient descent for log-linear models, and to SVM-Struct for max-margin models. The algorithms are applied to a multi-class problem as well as to a more complex large-scale parsing task. In all these settings, the EG algorithms presented here outperform the other methods.

Current and emerging screening methods to identify post-head-emergence frost adaptation in wheat and barley

Cereal crops can suffer substantial damage if frosts occur at heading. Identification of post-head-emergence frost (PHEF) resistance in cereals poses a number of unique and difficult challenges. Many decades of research have failed to identify genotypes with PHEF resistance that could offer economically significant benefit to growers. Research and breeding gains have been limited by the available screening systems. Using traditional frost screening systems, genotypes that escape frost injury in trials due to spatial temperature differences and/or small differences in phenology can be misidentified as resistant. We believe that by improving techniques to minimize frost escapes, such ofalse-positive' results can be confidently identified and eliminated. Artificial freezing chambers or manipulated natural frost treatments offer many potential advantages but are not yet at the stage where they can be reliably used for frost screening in breeding programmes. Here we describe the development of a novel photoperiod gradient method (PGM) that facilitates screening of genotypes of different phenology under natural field frosts at matched developmental stages. By identifying frost escapes and increasing the efficiency of field screening, the PGM ensures that research effort can be focused on finding genotypes with improved PHEF resistance. To maximize the likelihood of identifying PHEF resistance, we propose that the PGM form part of an integrated strategy to (i) source germplasm;(ii) facilitate high throughput screening; and (iii) permit detailed validation. PGM may also be useful in other studies where either a range of developmental stages and/or synchronized development are desired.

Development of a transient gradient enhanced non local continuum damage mechanics model for masonry

Due to the advent of varied types of masonry systems a comprehensive failure mechanism of masonry essential for the understanding of its behaviour is impossible to be determined from experimental testing. As masonry is predominantly used in wall structures a biaxial stress state dominates its failure mechanism. Biaxial testing will therefore be necessary for each type of masonry, which is expensive and time consuming. A computational method would be advantageous; however masonry is complex to model which requires advanced computational modelling methods. This thesis has formulated a damage mechanics inspired modelling method and has shown that the method effectively determines the failure mechanisms and deformation characteristics of masonry under biaxial states of loading.

An FETI-preconditioned conjuerate gradient method for large-scale stochastic finite element problems

In the spectral stochastic finite element method for analyzing an uncertain system. the uncertainty is represented by a set of random variables, and a quantity of Interest such as the system response is considered as a function of these random variables Consequently, the underlying Galerkin projection yields a block system of deterministic equations where the blocks are sparse but coupled. The solution of this algebraic system of equations becomes rapidly challenging when the size of the physical system and/or the level of uncertainty is increased This paper addresses this challenge by presenting a preconditioned conjugate gradient method for such block systems where the preconditioning step is based on the dual-primal finite element tearing and interconnecting method equipped with a Krylov subspace reusage technique for accelerating the iterative solution of systems with multiple and repeated right-hand sides. Preliminary performance results on a Linux Cluster suggest that the proposed Solution method is numerically scalable and demonstrate its potential for making the uncertainty quantification Of realistic systems tractable.

Strength design of composite beam using gradient and particle swarm optimization

This work addresses the optimum design of a composite box-beam structure subject to strength constraints. Such box-beams are used as the main load carrying members of helicopter rotor blades. A computationally efficient analytical model for box-beam is used. Optimal ply orientation angles are sought which maximize the failure margins with respect to the applied loading. The Tsai-Wu-Hahn failure criterion is used to calculate the reserve factor for each wall and ply and the minimum reserve factor is maximized. Ply angles are used as design variables and various cases of initial starting design and loadings are investigated. Both gradient-based and particle swarm optimization (PSO) methods are used. It is found that the optimization approach leads to the design of a box-beam with greatly improved reserve factors which can be useful for helicopter rotor structures. While the PSO yields globally best designs, the gradient-based method can also be used with appropriate starting designs to obtain useful designs efficiently. (C) 2006 Elsevier Ltd. All rights reserved.

Thermodynamic Study of Mixed Anionic Solid Solutions Using Gradient Solid Electrolytes System K2CO3 - K2SO4

The thermodynamic properties of K2CO3 -KSO, solid solutions with hexagonal structure have been measured using a solid-state cell, incorporating a composite solid electrolyte with step-changes in composition. The cell with the configuration Pt, CO2' + O2' || K2CO3 | K2(CO3)x(SO4)1-x || CO2'' + O2'' + Pt X =1 X=X was investigated in the temperature range of 925 to 1165 K. The composite gradient solid electrolyte consisted of pure K2CO3 at one extremity and the solid solution under study at the other. The Nernstian response of the cell to changes in partial pressures of CO2 and O2 at the electrodes and temperature was demonstrated. The activity of K2CO3 in the solid solution was measured by three techniques. All three methods gave identical results, indicating unit transport number for K+ ions and negligible diffusion potential due to concentration gradients of carbonate and sulfate ions. The activity of K2CO3 exhibits positive deviation from Raoult's law. The excess Gibbs energy of mixing of the solid solution can be represented using a subregular solution model DELTAG(E) = X(1 - X)[5030X + 4715(1 - X)] J mol-1 By combining this information with the phase diagram, mixing properties of the liquid phase were obtained.

Accelerated gradient based diffuse optical tomographic image reconstruction

Purpose: Fast reconstruction of interior optical parameter distribution using a new approach called Broyden-based model iterative image reconstruction (BMOBIIR) and adjoint Broyden-based MOBIIR (ABMOBIIR) of a tissue and a tissue mimicking phantom from boundary measurement data in diffuse optical tomography (DOT). Methods: DOT is a nonlinear and ill-posed inverse problem. Newton-based MOBIIR algorithm, which is generally used, requires repeated evaluation of the Jacobian which consumes bulk of the computation time for reconstruction. In this study, we propose a Broyden approach-based accelerated scheme for Jacobian computation and it is combined with conjugate gradient scheme (CGS) for fast reconstruction. The method makes explicit use of secant and adjoint information that can be obtained from forward solution of the diffusion equation. This approach reduces the computational time many fold by approximating the system Jacobian successively through low-rank updates. Results: Simulation studies have been carried out with single as well as multiple inhomogeneities. Algorithms are validated using an experimental study carried out on a pork tissue with fat acting as an inhomogeneity. The results obtained through the proposed BMOBIIR and ABMOBIIR approaches are compared with those of Newton-based MOBIIR algorithm. The mean squared error and execution time are used as metrics for comparing the results of reconstruction. Conclusions: We have shown through experimental and simulation studies that Broyden-based MOBIIR and adjoint Broyden-based methods are capable of reconstructing single as well as multiple inhomogeneities in tissue and a tissue-mimicking phantom. Broyden MOBIIR and adjoint Broyden MOBIIR methods are computationally simple and they result in much faster implementations because they avoid direct evaluation of Jacobian. The image reconstructions have been carried out with different initial values using Newton, Broyden, and adjoint Broyden approaches. These algorithms work well when the initial guess is close to the true solution. However, when initial guess is far away from true solution, Newton-based MOBIIR gives better reconstructed images. The proposed methods are found to be stable with noisy measurement data. (C) 2011 American Association of Physicists in Medicine. DOI: 10.1118/1.3531572]

Gradient solid electrolytes for thermodynamic measurements: System Na2CO3-Na2SO4

The thermodynamic properties of Na2CO3-Na2SO4 solid solution with hexagonal structure have been measured in the temperature range of 873 to 1073 K, using a composite-gradient solid electrolyte. The cell used can be represented as The composite-gradient solid electrolyte consisted of pure Na2CO3 at one extremity and the solid solution under study at the other, with variation in composition across the electrolyte. A CO2 + O2 + Ar gas mixture was used to fix the chemical potential of sodium at each electrode. The Nernstian response of the cell to changes in partial pressures of CO2 and O2 at the electrodes has been demonstrated. The activity of Na2CO3 in the solid solution was measured by two techniques. In the first method, the electromotive force (emf) of the cell was measured with the same CO2 + O2 + Ar mixture at both electrodes. The resultant emf is directly related to the activity of Na2CO3 at the solid solution electrode. By the second approach, the activity was calculated from the difference in compositions Of CO2 + O2 + Ar mixtures at the two electrodes required to produce a null emf. Both methods gave identical results. The second method is more suitable for gradient solid electrolytes that exhibit significant electronic conduction. The activity of Na2CO3 exhibits positive deviation from Raoult's law. The excess Gibbs' energy of mixing of the solid solution can be represented using a subregular solution model such as the following: DELTAG(E) = X(1 - X)[6500(+/-200)X + 3320(+/-80)(1 - X)J mol-1 where X is the mole fraction of Na2CO3. By combining this information with the phase diagram, mixing properties of the liquid phase are obtained.

Incremental natural-gradient actor-critic algorithms

We present four new reinforcement learning algorithms based on actor-critic and natural-gradient ideas, and provide their convergence proofs. Actor-critic rein- forcement learning methods are online approximations to policy iteration in which the value-function parameters are estimated using temporal difference learning and the policy parameters are updated by stochastic gradient descent. Methods based on policy gradients in this way are of special interest because of their com- patibility with function approximation methods, which are needed to handle large or infinite state spaces. The use of temporal difference learning in this way is of interest because in many applications it dramatically reduces the variance of the gradient estimates. The use of the natural gradient is of interest because it can produce better conditioned parameterizations and has been shown to further re- duce variance in some cases. Our results extend prior two-timescale convergence results for actor-critic methods by Konda and Tsitsiklis by using temporal differ- ence learning in the actor and by incorporating natural gradients, and they extend prior empirical studies of natural actor-critic methods by Peters, Vijayakumar and Schaal by providing the first convergence proofs and the first fully incremental algorithms.

Efficient Methods for Robust Classification Under Uncertainty in Kernel Matrices

In this paper we study the problem of designing SVM classifiers when the kernel matrix, K, is affected by uncertainty. Specifically K is modeled as a positive affine combination of given positive semi definite kernels, with the coefficients ranging in a norm-bounded uncertainty set. We treat the problem using the Robust Optimization methodology. This reduces the uncertain SVM problem into a deterministic conic quadratic problem which can be solved in principle by a polynomial time Interior Point (IP) algorithm. However, for large-scale classification problems, IP methods become intractable and one has to resort to first-order gradient type methods. The strategy we use here is to reformulate the robust counterpart of the uncertain SVM problem as a saddle point problem and employ a special gradient scheme which works directly on the convex-concave saddle function. The algorithm is a simplified version of a general scheme due to Juditski and Nemirovski (2011). It achieves an O(1/T-2) reduction of the initial error after T iterations. A comprehensive empirical study on both synthetic data and real-world protein structure data sets show that the proposed formulations achieve the desired robustness, and the saddle point based algorithm outperforms the IP method significantly.

Efficient gradient-free simplex method for estimation of optical properties in image-guided diffuse optical tomography

Typical image-guided diffuse optical tomographic image reconstruction procedures involve reduction of the number of optical parameters to be reconstructed equal to the number of distinct regions identified in the structural information provided by the traditional imaging modality. This makes the image reconstruction problem less ill-posed compared to traditional underdetermined cases. Still, the methods that are deployed in this case are same as those used for traditional diffuse optical image reconstruction, which involves a regularization term as well as computation of the Jacobian. A gradient-free Nelder-Mead simplex method is proposed here to perform the image reconstruction procedure and is shown to provide solutions that closely match ones obtained using established methods, even in highly noisy data. The proposed method also has the distinct advantage of being more efficient owing to being regularization free, involving only repeated forward calculations. (C) 2013 Society of Photo-Optical Instrumentation Engineers (SPIE)

Study of structures, energies and vibrational frequencies of (O-2)(n)(+) (n=2-5) clusters by GGA and meta-GGA density functional methods

Using Generalized Gradient Approximation (GGA) and meta-GGA density functional methods, structures, binding energies and harmonic vibrational frequencies for the clusters O-4(+), O-6(+), O-8(+) and O-10(+) have been calculated. The stable structures of O-4(+), O-6(+), O-8(+) and O-10(+) have point groups D-2h, D-3h, D-4h, and D-5h optimized on the quartet, sextet, octet and dectet potential energy surfaces, respectively. Rectangular (D-2h) O-4(+) has been found to be more stable compared to trans-planar (C-2h) on the quartet potential energy surface. Cyclic structure (D-3h) of CA cluster ion has been calculated to be more stable than other structures. Binding energy (B.E.) of the cyclic O-6(+) is in good agreement with experimental measurement. The zero-point corrected B.E. of O-8(+) with D4h symmetry on the octet potential energy surface and zero-point corrected B.E. of O-10(+) with D-5h symmetry on the dectet potential energy surface are also in good agreement with experimental values. The B.E. value for O-4(+) is close to the experimental value when single point energy is calculated by Brueckner coupled-cluster method, BD(T). (C) 2014 Elsevier B.V. All rights reserved.

A Smooth Discretization Bridging Finite Element and Mesh-free Methods Using Polynomial Reproducing Simplex Splines

This work sets forth a `hybrid' discretization scheme utilizing bivariate simplex splines as kernels in a polynomial reproducing scheme constructed over a conventional Finite Element Method (FEM)-like domain discretization based on Delaunay triangulation. Careful construction of the simplex spline knotset ensures the success of the polynomial reproduction procedure at all points in the domain of interest, a significant advancement over its precursor, the DMS-FEM. The shape functions in the proposed method inherit the global continuity (Cp-1) and local supports of the simplex splines of degree p. In the proposed scheme, the triangles comprising the domain discretization also serve as background cells for numerical integration which here are near-aligned to the supports of the shape functions (and their intersections), thus considerably ameliorating an oft-cited source of inaccuracy in the numerical integration of mesh-free (MF) schemes. Numerical experiments show the proposed method requires lower order quadrature rules for accurate evaluation of integrals in the Galerkin weak form. Numerical demonstrations of optimal convergence rates for a few test cases are given and the method is also implemented to compute crack-tip fields in a gradient-enhanced elasticity model.

Simultaneous perturbation methods for adaptive labor staffing in service systems

We consider the problem of optimizing the workforce of a service system. Adapting the staffing levels in such systems is non-trivial due to large variations in workload and the large number of system parameters do not allow for a brute force search. Further, because these parameters change on a weekly basis, the optimization should not take longer than a few hours. Our aim is to find the optimum staffing levels from a discrete high-dimensional parameter set, that minimizes the long run average of the single-stage cost function, while adhering to the constraints relating to queue stability and service-level agreement (SLA) compliance. The single-stage cost function balances the conflicting objectives of utilizing workers better and attaining the target SLAs. We formulate this problem as a constrained parameterized Markov cost process parameterized by the (discrete) staffing levels. We propose novel simultaneous perturbation stochastic approximation (SPSA)-based algorithms for solving the above problem. The algorithms include both first-order as well as second-order methods and incorporate SPSA-based gradient/Hessian estimates for primal descent, while performing dual ascent for the Lagrange multipliers. Both algorithms are online and update the staffing levels in an incremental fashion. Further, they involve a certain generalized smooth projection operator, which is essential to project the continuous-valued worker parameter tuned by our algorithms onto the discrete set. The smoothness is necessary to ensure that the underlying transition dynamics of the constrained Markov cost process is itself smooth (as a function of the continuous-valued parameter): a critical requirement to prove the convergence of both algorithms. We validate our algorithms via performance simulations based on data from five real-life service systems. For the sake of comparison, we also implement a scatter search based algorithm using state-of-the-art optimization tool-kit OptQuest. From the experiments, we observe that both our algorithms converge empirically and consistently outperform OptQuest in most of the settings considered. This finding coupled with the computational advantage of our algorithms make them amenable for adaptive labor staffing in real-life service systems.

Gradient echo single scan inversion recovery: application to proton and fluorine relaxation studies

Single scan longitudinal relaxation measurement experiments enable rapid estimation of the spin-lattice relaxation time (T-1) as the time series of spin relaxation is encoded spatially in the sample at different slices resulting in an order of magnitude saving in time. We consider here a single scan inversion recovery pulse sequence that incorporates a gradient echo sequence. The proposed pulse sequence provides spectra with significantly enhanced signal to noise ratio leading to an accurate estimation of T-1 values. The method is applicable for measuring a range of T-1 values, thus indicating the possibility of routine use of the method for several systems. A comparative study of different single scan methods currently available is presented, and the advantage of the proposed sequence is highlighted. The possibility of the use of the method for the study of cross-correlation effects for the case of fluorine in a single shot is also demonstrated. Copyright (C) 2015 John Wiley & Sons, Ltd.