Showing papers on "Maxima and minima published in 2014"

Continuous Energy Minimization for Multitarget Tracking

Abstract: Many recent advances in multiple target tracking aim at finding a (nearly) optimal set of trajectories within a temporal window. To handle the large space of possible trajectory hypotheses, it is typically reduced to a finite set by some form of data-driven or regular discretization. In this work, we propose an alternative formulation of multitarget tracking as minimization of a continuous energy. Contrary to recent approaches, we focus on designing an energy that corresponds to a more complete representation of the problem, rather than one that is amenable to global optimization. Besides the image evidence, the energy function takes into account physical constraints, such as target dynamics, mutual exclusion, and track persistence. In addition, partial image evidence is handled with explicit occlusion reasoning, and different targets are disambiguated with an appearance model. To nevertheless find strong local minima of the proposed nonconvex energy, we construct a suitable optimization scheme that alternates between continuous conjugate gradient descent and discrete transdimensional jump moves. These moves, which are executed such that they always reduce the energy, allow the search to escape weak minima and explore a much larger portion of the search space of varying dimensionality. We demonstrate the validity of our approach with an extensive quantitative evaluation on several public data sets.

Identifying and attacking the saddle point problem in high-dimensional non-convex optimization

Abstract: A central challenge to many fields of science and engineering involves minimizing non-convex error functions over continuous, high dimensional spaces. Gradient descent or quasi-Newton methods are almost ubiquitously used to perform such minimizations, and it is often thought that a main source of difficulty for these local methods to find the global minimum is the proliferation of local minima with much higher error than the global minimum. Here we argue, based on results from statistical physics, random matrix theory, neural network theory, and empirical evidence, that a deeper and more profound difficulty originates from the proliferation of saddle points, not local minima, especially in high dimensional problems of practical interest. Such saddle points are surrounded by high error plateaus that can dramatically slow down learning, and give the illusory impression of the existence of a local minimum. Motivated by these arguments, we propose a new approach to second-order optimization, the saddle-free Newton method, that can rapidly escape high dimensional saddle points, unlike gradient descent and quasi-Newton methods. We apply this algorithm to deep or recurrent neural network training, and provide numerical evidence for its superior optimization performance.

Identifying and attacking the saddle point problem in high-dimensional non-convex optimization

Abstract: A central challenge to many fields of science and engineering involves minimizing non-convex error functions over continuous, high dimensional spaces. Gradient descent or quasi-Newton methods are almost ubiquitously used to perform such minimizations, and it is often thought that a main source of difficulty for these local methods to find the global minimum is the proliferation of local minima with much higher error than the global minimum. Here we argue, based on results from statistical physics, random matrix theory, neural network theory, and empirical evidence, that a deeper and more profound difficulty originates from the proliferation of saddle points, not local minima, especially in high dimensional problems of practical interest. Such saddle points are surrounded by high error plateaus that can dramatically slow down learning, and give the illusory impression of the existence of a local minimum. Motivated by these arguments, we propose a new approach to second-order optimization, the saddle-free Newton method, that can rapidly escape high dimensional saddle points, unlike gradient descent and quasi-Newton methods. We apply this algorithm to deep or recurrent neural network training, and provide numerical evidence for its superior optimization performance.

The Loss Surfaces of Multilayer Networks

Abstract: We study the connection between the highly non-convex loss function of a simple model of the fully-connected feed-forward neural network and the Hamiltonian of the spherical spin-glass model under the assumptions of: i) variable independence, ii) redundancy in network parametrization, and iii) uniformity. These assumptions enable us to explain the complexity of the fully decoupled neural network through the prism of the results from random matrix theory. We show that for large-size decoupled networks the lowest critical values of the random loss function form a layered structure and they are located in a well-defined band lower-bounded by the global minimum. The number of local minima outside that band diminishes exponentially with the size of the network. We empirically verify that the mathematical model exhibits similar behavior as the computer simulations, despite the presence of high dependencies in real networks. We conjecture that both simulated annealing and SGD converge to the band of low critical points, and that all critical points found there are local minima of high quality measured by the test error. This emphasizes a major difference between large- and small-size networks where for the latter poor quality local minima have non-zero probability of being recovered. Finally, we prove that recovering the global minimum becomes harder as the network size increases and that it is in practice irrelevant as global minimum often leads to overfitting.

MUSIC for Single-Snapshot Spectral Estimation: Stability and Super-resolution

Abstract: This paper studies the problem of line spectral estimation in the continuum of a bounded interval with one snapshot of array measurement. The single-snapshot measurement data is turned into a Hankel data matrix which admits the Vandermonde decomposition and is suitable for the MUSIC algorithm. The MUSIC algorithm amounts to finding the null space (the noise space) of the Hankel matrix, forming the noise-space correlation function and identifying the s smallest local minima of the noise-space correlation as the frequency set. In the noise-free case exact reconstruction is guaranteed for any arbitrary set of frequencies as long as the number of measurements is at least twice the number of distinct frequencies to be recovered. In the presence of noise the stability analysis shows that the perturbation of the noise-space correlation is proportional to the spectral norm of the noise matrix as long as the latter is smaller than the smallest (nonzero) singular value of the noiseless Hankel data matrix. Under the assumption that frequencies are separated by at least twice the Rayleigh Length (RL), the stability of the noise-space correlation is proved by means of novel discrete Ingham inequalities which provide bounds on nonzero singular values of the noiseless Hankel data matrix. The numerical performance of MUSIC is tested in comparison with other algorithms such as BLO-OMP and SDP (TV-min). While BLO-OMP is the stablest algorithm for frequencies separated above 4 RL, MUSIC becomes the best performing one for frequencies separated between 2 RL and 3 RL. Also, MUSIC is more efficient than other methods. MUSIC truly shines when the frequency separation drops to 1 RL or below when all other methods fail. Indeed, the resolution length of MUSIC decreases to zero as noise decreases to zero as a power law with an exponent much smaller than an upper bound established by Donoho.

Globally-biased Disimpl algorithm for expensive global optimization

Abstract: Direct-type global optimization algorithms often spend an excessive number of function evaluations on problems with many local optima exploring suboptimal local minima, thereby delaying discovery of the global minimum. In this paper, a globally-biased simplicial partition Disimpl algorithm for global optimization of expensive Lipschitz continuous functions with an unknown Lipschitz constant is proposed. A scheme for an adaptive balancing of local and global information during the search is introduced, implemented, experimentally investigated, and compared with the well-known Direct and Direct l methods. Extensive numerical experiments executed on 800 multidimensional multiextremal test functions show a promising performance of the new acceleration technique with respect to competitors.

Approximate Symmetry Detection in Partial 3D Meshes

Abstract: Symmetry is a common characteristic in natural and man-made objects. Its ubiquitous nature can be exploited to facilitate the analysis and processing of computational representations of real objects. In particular, in computer graphics, the detection of symmetries in 3D geometry has enabled a number of applications in modeling and reconstruction. However, the problem of symmetry detection in incomplete geometry remains a challenging task. In this paper, we propose a vote-based approach to detect symmetry in 3D shapes, with special interest in models with large missing parts. Our algorithm generates a set of candidate symmetries by matching local maxima of a surface function based on the heat diffusion in local domains, which guarantee robustness to missing data. In order to deal with local perturbations, we propose a multi-scale surface function that is useful to select a set of distinctive points over which the approximate symmetries are defined. In addition, we introduce a vote-based scheme that is aware of the partiality, and therefore reduces the number of false positive votes for the candidate symmetries. We show the effectiveness of our method in a varied set of 3D shapes and different levels of partiality. Furthermore, we show the applicability of our algorithm in the repair and completion of challenging reassembled objects in the context of cultural heritage.

On the saddle point problem for non-convex optimization

Abstract: A central challenge to many fields of science and engineering involves minimizing non-convex error functions over continuous, high dimensional spaces. Gradient descent or quasi-Newton methods are almost ubiquitously used to perform such minimizations, and it is often thought that a main source of difficulty for the ability of these local methods to find the global minimum is the proliferation of local minima with much higher error than the global minimum. Here we argue, based on results from statistical physics, random matrix theory, and neural network theory, that a deeper and more profound difficulty originates from the proliferation of saddle points, not local minima, especially in high dimensional problems of practical interest. Such saddle points are surrounded by high error plateaus that can dramatically slow down learning, and give the illusory impression of the existence of a local minimum. Motivated by these arguments, we propose a new algorithm, the saddle-free Newton method, that can rapidly escape high dimensional saddle points, unlike gradient descent and quasi-Newton methods. We apply this algorithm to deep neural network training, and provide preliminary numerical evidence for its superior performance.

A fast and accurate approximation for planar pose graph optimization

Abstract: This work investigates the pose graph optimization problem, which arises in maximum likelihood approaches to simultaneous localization and mapping SLAM. State-of-the-art approaches have been demonstrated to be very efficient in medium- and large-sized scenarios; however, their convergence to the maximum likelihood estimate heavily relies on the quality of the initial guess. We show that, in planar scenarios, pose graph optimization has a very peculiar structure. The problem of estimating robot orientations from relative orientation measurements is a quadratic optimization problem after computing suitable regularization terms; moreover, given robot orientations, the overall optimization problem becomes quadratic. We exploit these observations to design an approximation of the maximum likelihood estimate, which does not require the availability of an initial guess. The approximation, named LAGO Linear Approximation for pose Graph Optimization, can be used as a stand-alone tool or can bootstrap state-of-the-art techniques, reducing the risk of being trapped in local minima. We provide analytical results on existence and sub-optimality of LAGO, and we discuss the factors influencing its quality. Experimental results demonstrate that LAGO is accurate in common SLAM problems. Moreover, it is remarkably faster than state-of-the-art techniques, and is able to solve very large-scale problems in a few seconds.

Noisy Gradient Descent Bit-Flip Decoding for LDPC Codes

Abstract: A modified Gradient Descent Bit Flipping (GDBF) algorithm is proposed for decoding Low Density Parity Check (LDPC) codes on the binary-input additive white Gaussian noise channel. The new algorithm, called Noisy GDBF (NGDBF), introduces a random perturbation into each symbol metric at each iteration. The noise perturbation allows the algorithm to escape from undesirable local maxima, resulting in improved performance. A combination of heuristic improvements to the algorithm are proposed and evaluated. When the proposed heuristics are applied, NGDBF performs better than any previously reported GDBF variant, and comes within 0.5 dB of the belief propagation algorithm for several tested codes. Unlike other previous GDBF algorithms that provide an escape from local maxima, the proposed algorithm uses only local, fully parallelizable operations and does not require computing a global objective function or a sort over symbol metrics, making it highly efficient in comparison. The proposed NGDBF algorithm requires channel state information which must be obtained from a signal to noise ratio (SNR) estimator. Architectural details are presented for implementing the NGDBF algorithm. Complexity analysis and optimizations are also discussed.

Energy landscapes of resting-state brain networks

Abstract: During rest, the human brain performs essential functions such as memory maintenance, which are associated with resting-state brain networks (RSNs) including the default-mode network (DMN) and frontoparietal network (FPN). Previous studies based on spiking-neuron network models and their reduced models, as well as those based on imaging data, suggest that resting-state network activity can be captured as attractor dynamics, i.e., dynamics of the brain state toward an attractive state and transitions between different attractors. Here, we analyze the energy landscapes of the RSNs by applying the maximum entropy model, or equivalently the Ising spin model, to human RSN data. We use the previously estimated parameter values to define the energy landscape, and the disconnectivity graph method to estimate the number of local energy minima (equivalent to attractors in attractor dynamics), the basin size, and hierarchical relationships among the different local minima. In both of the DMN and FPN, low-energy local minima tended to have large basins. A majority of the network states belonged to a basin of one of a few local minima. Therefore, a small number of local minima constituted the backbone of each RSN. In the DMN, the energy landscape consisted of two groups of low-energy local minima that are separated by a relatively high energy barrier. Within each group, the activity patterns of the local minima were similar, and different minima were connected by relatively low energy barriers. In the FPN, all dominant local minima were separated by relatively low energy barriers such that they formed a single coarse-grained global minimum. Our results indicate that multistable attractor dynamics may underlie the DMN, but not the FPN, and assist memory maintenance with different memory states.

Alternative waveform inversion for surface wave analysis in 2-D media

Abstract: In the context of near surface seismic imaging (a few hundreds of metres), we propose an alternative approach for inversion of surface waves in 2-D media with laterally varying velocities. It is based on Full Waveform Inversion (FWI) but using an alternative objective function formulated in the frequency–wavenumber f − k domain. The classical FWI objective function suffers from severe local minima problems in the presence of surface waves. It thus requires a very accurate initial model. The proposed objective function is similar to the one used in classical surface wave analysis. In this approach, the data are first split using sliding windows in the time–space t − x domain. For each window, the amplitude of the f − k spectrum is computed. The objective function measures the least-squares misfit between the amplitude of observed and modelled 2-D Fourier transformed data sets. We call this formulation the windowed-amplitude waveform inversion (w-AWI). The w-AWI objective function reduces some local minima problems as shown here through numerical examples. The global minimum basin is wider in the w-AWI approach than in FWI. Synthetic examples show that w-AWI may achieve convergence if the lowest data frequency content is twice higher than the one needed by FWI. For elastic inversion, w-AWI can be used to reconstruct a velocity model explaining surface waves. This surface wave inversion procedure can be used to retrieve near-surface model parameters in lateral-varying media

Topology Trivialization and Large Deviations for the Minimum in the Simplest Random Optimization

Abstract: Finding the global minimum of a cost function given by the sum of a quadratic and a linear form in N real variables over (N−1)-dimensional sphere is one of the simplest, yet paradigmatic problems in Optimization Theory known as the “trust region subproblem” or “constraint least square problem”. When both terms in the cost function are random this amounts to studying the ground state energy of the simplest spherical spin glass in a random magnetic field. We first identify and study two distinct large-N scaling regimes in which the linear term (magnetic field) leads to a gradual topology trivialization, i.e. reduction in the total number $\mathcal{N}_{tot}$ of critical (stationary) points in the cost function landscape. In the first regime $\mathcal{N}_{tot}$ remains of the order N and the cost function (energy) has generically two almost degenerate minima with the Tracy-Widom (TW) statistics. In the second regime the number of critical points is of the order of unity with a finite probability for a single minimum. In that case the mean total number of extrema (minima and maxima) of the cost function is given by the Laplace transform of the TW density, and the distribution of the global minimum energy is expected to take a universal scaling form generalizing the TW law. Though the full form of that distribution is not yet known to us, one of its far tails can be inferred from the large deviation theory for the global minimum. In the rest of the paper we show how to use the replica method to obtain the probability density of the minimum energy in the large-deviation approximation by finding both the rate function and the leading pre-exponential factor.

gDLS: A Scalable Solution to the Generalized Pose and Scale Problem

Abstract: In this work, we present a scalable least-squares solution for computing a seven degree-of-freedom similarity transform. Our method utilizes the generalized camera model to compute relative rotation, translation, and scale from four or more 2D-3D correspondences. In particular, structure and motion estimations from monocular cameras lack scale without specific calibration. As such, our methods have applications in loop closure in visual odometry and registering multiple structure from motion reconstructions where scale must be recovered. We formulate the generalized pose and scale problem as a minimization of a least squares cost function and solve this minimization without iterations or initialization. Additionally, we obtain all minima of the cost function. The order of the polynomial system that we solve is independent of the number of points, allowing our overall approach to scale favorably. We evaluate our method experimentally on synthetic and real datasets and demonstrate that our methods produce higher accuracy similarity transform solutions than existing methods.

Genetic optimization of neural network and fuzzy logic for oil bubble point pressure modeling

Abstract: Bubble point pressure is a critical pressure-volume-temperature (PVT) property of reservoir fluid, which plays an important role in almost all tasks involved in reservoir and production engineering. We developed two sophisticated models to estimate bubble point pressure from gas specific gravity, oil gravity, solution gas oil ratio, and reservoir temperature. Neural network and adaptive neuro-fuzzy inference system are powerful tools for extracting the underlying dependency of a set of input/output data. However, the mentioned tools are in danger of sticking in local minima. The present study went further by optimizing fuzzy logic and neural network models using the genetic algorithm in charge of eliminating the risk of being exposed to local minima. This strategy is capable of significantly improving the accuracy of both neural network and fuzzy logic models. The proposed methodology was successfully applied to a dataset of 153 PVT data points. Results showed that the genetic algorithm can serve the neural network and neurofuzzy models from local minima trapping, which might occur through back-propagation algorithm.

Complex network approach to characterize the statistical features of the sunspot series

Abstract: Complex network approaches have been recently developed as an alternative framework to study the statistical features of time-series data. We perform a visibility-graph analysis on both the daily and monthly sunspot series. Based on the data, we propose two ways to construct the network: one is from the original observable measurements and the other is from a negative-inverse-transformed series. The degree distribution of the derived networks for the strong maxima has clear non-Gaussian properties, while the degree distribution for minima is bimodal. The long-term variation of the cycles is reflected by hubs in the network that span relatively large time intervals. Based on standard network structural measures, we propose to characterize the long-term correlations by waiting times between two subsequent events. The persistence range of the solar cycles has been identified over 15‐1000 days by a power-law regime with scaling exponent = 2.04 of the occurrence time of two subsequent strong minima. In contrast, a persistent trend is not present in the maximal numbers, although maxima do have significant deviations from an exponential form. Our results suggest some new insights for evaluating existing models.

Distribution Loss Minimization With Guaranteed Error Bound

Abstract: Determining loss minimum configuration in a distribution network is a hard discrete optimization problem involving many variables. Since more and more dispersed generators are installed on the demand side of power systems and they are reconfigured frequently, developing automatic approaches is indispensable for effectively managing a large-scale distribution network. Existing fast methods employ local updates that gradually improve the loss to solve such an optimization problem. However, they eventually get stuck at local minima, resulting in arbitrarily poor results. In contrast, this paper presents a novel optimization method that provides an error bound on the solution quality. Thus, the obtained solution quality can be evaluated in comparison to the global optimal solution. Instead of using local updates, we construct a highly compressed search space using a binary decision diagram and reduce the optimization problem to a shortest path-finding problem. Our method was shown to be not only accurate but also remarkably efficient; optimization of a large-scale model network with 468 switches was solved in three hours with 1.56% relative error bound.

Online generation of homotopically distinct navigation paths.

Abstract: In mobile robot navigation, cost functions are a popular approach to generate feasible, safe paths that avoid obstacles and that allow the robot to get from its starting position to the goal position. Alternative ways to navigate around the obstacles typically correspond to different local minima in the cost function. In this paper we present a highly effective approach to overcome such local minima and to quickly propose a set of alternative, topologically different and optimized paths. We furthermore describe how to maintain a set of optimized trajectory alternatives to reduce optimization efforts when the robot has to adapt to changes in the environment. We demonstrate in experiments that our method outperforms a stateof-the-art approach by an order of magnitude in computation time, which allows a robot to use our method online during navigation. We furthermore demonstrate that the approach of using a set of qualitatively different trajectories is beneficial in shared autonomy settings, where a user operating a wheelchair can quickly switch between topologically different trajectories.

Exploring the topography of the stress-modified energy landscapes of mechanosensitive molecules.

Abstract: We propose a method for computing the activation barrier for chemical reactions involving molecules subjected to mechanical stress. The method avoids reactant and transition-state saddle optimizations at every force by, instead, solving the differential equations governing the force dependence of the critical points (i.e., minima and saddles) on the system's potential energy surface (PES). As a result, only zero-force geometry optimization (or, more generally, optimization performed at a single force value) is required by the method. In many cases, minima and transition-state saddles only exist within a range of forces and disappear beyond a certain critical point. Our method identifies such force-induced instabilities as points at which one of the Hessian eigenvalues vanishes. We elucidate the nature of those instabilities as fold and cusp catastrophes, where two or three critical points on the force-modified PES coalesce, and provide a classification of various physically distinct instability scenarios, each illustrated with a concrete chemical example.

On using extreme values to detect global stability thresholds in multi-stable systems: The case of transitional plane Couette flow

Abstract: Extreme Value Theory (EVT) is exploited to determine the global stability threshold R g of plane Couette flow – the flow of a viscous fluid in the space between two parallel plates – whose laminar or turbulent behavior depends on the Reynolds number R. Even if the existence of a global stability threshold has been detected in simulations and experiments, its numerical value has not been unequivocally defined. R g is the value such that for R > R g , turbulence is sustained, whereas for R R g it is transient and eventually decays. We address the problem of determining R g by using the extremes – maxima and minima – of the perturbation energy fluctuations. When R ≫ R g , both the positive and negative extremes are bounded. As the critical Reynolds number is approached from above, the probability of observing a very low minimum increases causing asymmetries in the distributions of maxima and minima. On the other hand, the maxima distribution is unaffected as the fluctuations towards higher values of the perturbation energy remain bounded. This tipping point can be detected by fitting the data to the Generalized Extreme Value (GEV) distribution and by identifying R g as the value of R such that the shape parameter of the GEV for the minima changes sign from negative to positive. The results are supported by the analysis of theoretical models which feature a bistable behavior.

cluster: Searching for Unique Low Energy Minima of Structures Using a Novel Implementation of a Genetic Algorithm

Abstract: A new flexible implementation of a genetic algorithm for locating unique low energy minima of isomers of clusters is described and tested. The strategy employed can be applied to molecular or atomic clusters and has a flexible input structure so that a system with several different elements can be built up from a set of individual atoms or from fragments made up of groups of atoms. This cluster program is tested on several systems, and the results are compared to computational and experimental data from previous studies. The quality of the algorithm for locating reliably the most competitive low energy structures of an assembly of atoms is examined for strongly bound Si–Li clusters, and ZnF2 clusters, and the more weakly interacting water trimers. The use of the nuclear repulsion energy as a duplication criterion, an increasing population size, and avoiding mutation steps without loss of efficacy are distinguishing features of the program. For the Si–Li clusters, a few new low energy minima are identified...

Explorations on high dimensional landscapes

Abstract: Finding minima of a real valued non-convex function over a high dimensional space is a major challenge in science. We provide evidence that some such functions that are defined on high dimensional domains have a narrow band of values whose pre-image contains the bulk of its critical points. This is in contrast with the low dimensional picture in which this band is wide. Our simulations agree with the previous theoretical work on spin glasses that proves the existence of such a band when the dimension of the domain tends to infinity. Furthermore our experiments on teacher-student networks with the MNIST dataset establish a similar phenomenon in deep networks. We finally observe that both the gradient descent and the stochastic gradient descent methods can reach this level within the same number of steps.

Extending DFT-based genetic algorithms by atom-to-place re-assignment via perturbation theory: A systematic and unbiased approach to structures of mixed-metallic clusters

Abstract: Energy surfaces of metal clusters usually show a large variety of local minima For homo-metallic species the energetically lowest can be found reliably with genetic algorithms, in combination with density functional theory without system-specific parameters For mixed-metallic clusters this is much more difficult, as for a given arrangement of nuclei one has to find additionally the best of many possibilities of assigning different metal types to the individual positions In the framework of electronic structure methods this second issue is treatable at comparably low cost at least for elements with similar atomic number by means of first-order perturbation theory, as shown previously [F Weigend, C Schrodt, and R Ahlrichs, J Chem Phys 121, 10380 (2004)] In the present contribution the extension of a genetic algorithm with the re-assignment of atom types to atom sites is proposed and tested for the search of the global minima of PtHf12 and [LaPb7Bi7]4− For both cases the (putative) global minimum

Minima hopping guided path search: An efficient method for finding complex chemical reaction pathways

Abstract: The Minima Hopping global optimization method uses physically realizable molecular dynamics moves in combination with an energy feedback that guarantees the escape from any potential energy funnel. For the purpose of finding reaction pathways, we argue that Minima Hopping is particularly suitable as a guide through the potential energy landscape and as a generator for pairs of minima that can be used as input structures for methods capable of finding transition states between two minima. For Lennard-Jones benchmark systems we compared this Minima Hopping guided path search method to a known approach for the exploration of potential energy landscapes that is based on deterministic mode-following. Although we used a stabilized mode-following technique that reliably allows to follow distinct directions when escaping from a local minimum, we observed that Minima Hopping guided path search is far superior in finding lowest-barrier reaction pathways. We, therefore, suggest that Minima Hopping guided path search can be used as a simple and efficient way to identify energetically low-lying chemical reaction pathways. Finally, we applied the Minima Hopping guided path search approach to 75-atom and 102-atom Lennard-Jones systems. For the 75-atom system we found pathways whose highest energies are significantly lower than the highest energy along the previously published lowest-barrier pathway. Furthermore, many of these pathways contain a smaller number of intermediate transition states than the previously publish lowest-barrier pathway. In case of the 102-atom system Minima Hopping guided path search found a previously unknown and energetically low-lying funnel.

Minima Hopping Guided Path Search: An Efficient Method for Finding Complex Chemical Reaction Pathways

Abstract: The Minima Hopping global optimization method uses physically realizable molecular dynamics moves in combination with an energy feedback that guarantees the escape from any potential energy funnel. For the purpose of finding reactions pathways, we argue that Minima Hopping is particularly suitable as a guide through the potential energy landscape and as a generator for pairs of minima that can be used as input structures for methods capable of finding transition states between two minima. For Lennard-Jones benchmark systems we compared this Minima Hopping guided path search method to a known approach for the exploration of potential energy landscapes that is based on deterministic mode-following. Although we used a stabilized mode-following technique that reliably allows to follow distinct directions when escaping from a local minimum, we observed that Minima Hopping guided path search is far superior in finding lowest-barrier reaction pathways. We therefore suggest that Minima Hopping guided path search can be used as a simple and efficient way to identify energetically low-lying chemical reaction pathways. Finally we applied the Minima Hopping guided path search approach to 75-atom and 102-atom Lennard Jones systems. For the 75-atom system we found pathways whose highest energies are significantly lower than the highest energy along the previously published lowest-barrier pathway. Furthermore, many of these pathways contain a smaller number of intermediate transition states than the previously publish lowest-barrier pathway. In case of the 102-atom system Minima Hopping guided path search found a previously unknown and energetically low-lying funnel.

Identification time-delayed fractional order chaos with functional extrema model via differential evolution

Abstract: In this paper, a novel inversion mechanism of functional extrema model via the differential evolution algorithms (DE), is proposed to exactly identify time-delays fractional order chaos systems. With the functional extrema model, the unknown time-delays, systematic parameters and fractional-orders of the fractional chaos, are converted into independent variables of a non-negative multiple modal functions' minimization, as a particular case of the functional extrema model's minimization. And the objective of the model is to find their optimal combinations by DE in the predefined intervals, such that the objective functional is minimized. Simulations are done to identify two classical time-delayed fractional chaos, Logistic and Chen system, both in cases with noise and without. The experiments' results show that the proposed inversion mechanism for time-delay fractional-order chaotic systems is a successful methods, with the advantages of high precision and robustness.

A model reduction approach to numerical inversion for a parabolic partial differential equation

Abstract: We propose a novel numerical inversion algorithm for the coefficients of parabolic partial differential equations, based on model reduction. The study is motivated by the application of controlled source electromagnetic exploration, where the unknown is the subsurface electrical resistivity and the data are time resolved surface measurements of the magneticfield. The algorithm presented in this paper considers inversion in one and two dimensions. The reduced model is obtained with rational interpolation in the frequency (Laplace) domain and a rational Krylov subspace projection method. It amounts to a nonlinear mapping from the function space of the unknown resistivity to the small dimensional space of the parameters of the reduced model. We use this mapping as a nonlinear preconditioner for the Gauss‐Newton iterative solution of the inverse problem. The advantage of the inversion algorithm is twofold. First, the nonlinear preconditioner resolves most of the nonlinearity of the problem. Thus the iterations are less likely to get stuck in local minima and the convergence is fast. Second, the inversion is computationally efficient because it avoids repeated accurate simulations of the time-domain response. We study the stability of the inversion algorithm for various rational Krylov subspaces, and assess its performance with numerical experiments.