Downloads & Free Reading Options - Results

Today's edition of Ben's Story of the Day features:1) People were protesting outside of Facebook's offices in London in the hopes of changing the way they moderate pictures of nipples vs tattoos of nipples on breast cancer survivors. Though some protestors dressed up as giant breasts, hopefully no one acted like one.2) Gender reveals have caused much damage and devastation around the world. An expectant father from Tennessee got the good news and simply had to fire off a few celebratory rounds from his shotgun into the air prompting 3 nearby schools to go into lockdown.3) Some collectors have extremely detailed and expensive collections of just about everything you can think of. One such collection was tragically thrown out by a man's parents after he moved out. A lawsuit was filed and won by the 43-year-old man whose parents now have to pay him $30,0441 to cover for the collection of pornography and sex toys.Which one will be crowned Ben's Story of the Day? Which one is your story of the day?

Twitter's Algorithm, the oldest uploaded version: taken from https://github.com/twitter/the-algorithm

We consider the numerical solution of the equation - \Delta u - f(u) = g, for the unknown u satisfying Dirichlet conditions in a bounded domain. The nonlinearity f has bounded, continuous derivative. The algorithm uses the finite element method combined with a global Lyapunov-Schmidt decomposition.

A transfer-matrix algorithm is presented herein as a beginning to study the transmission characteristics of coherent light through three-dimensional periodic microstructures, in which the structures are treated as two-dimensional-layer stacks and multiple reflections are considered negligible. The spatial-correlated noise is further introduced layer by layer to realize the actual decoherence of the light and allows for statistical investigation of the partial spatially coherent optics in transparent mediums. Numerical analyses show comparable results to the Gaussian Schell model in free-space cases, indicating the validity of the algorithms.

Herman's algorithm is a synchronous randomized protocol for achieving self-stabilization in a token ring consisting of N processes. The interaction of tokens makes the dynamics of the protocol very difficult to analyze. In this paper we study the expected time to stabilization in terms of the initial configuration. It is straightforward that the algorithm achieves stabilization almost surely from any initial configuration, and it is known that the worst-case expected time to stabilization (with respect to the initial configuration) is Theta(N^2). Our first contribution is to give an upper bound of 0.64 N^2 on the expected stabilization time, improving on previous upper bounds and reducing the gap with the best existing lower bound. We also introduce an asynchronous version of the protocol, showing a similar O(N^2) convergence bound in this case. Assuming that errors arise from the corruption of some number k of bits, where k is fixed independently of the size of the ring, we show that the expected time to stabilization is O(N). This reveals a hitherto unknown and highly desirable property of Herman's algorithm: it recovers quickly from bounded errors. We also show that if the initial configuration arises by resetting each bit independently and uniformly at random, then stabilization is significantly faster than in the worst case.

A low storage algorithm for constructing isogenies between ordinary elliptic curves was proposed by Galbraith, Hess and Smart (GHS). We give an improvement of this algorithm by modifying the pseudorandom walk so that lower-degree isogenies are used more frequently. This is motivated by the fact that high degree isogenies are slower to compute than low degree ones. We analyse the running time of the parallel collision search algorithm when the partitioning is uneven. We also give experimental results. We conclude that our algorithm is around 14 times faster than the GHS algorithm when constructing horizontal isogenies between random isogenous elliptic curves over a 160-bit prime field. The results apply to generic adding walks and the more general group action inverse problem; a speed-up is obtained whenever the cost of computing edges in the graph varies significantly.

The paper proposes a numerically stable recursive algorithm for the exact computation of the linear-chain conditional random field gradient. It operates as a forward algorithm over the log-domain expectation semiring and has the purpose of enhancing memory efficiency when applied to long observation sequences. Unlike the traditional algorithm based on the forward-backward recursions, the memory complexity of our algorithm does not depend on the sequence length. The experiments on real data show that it can be useful for the problems which deal with long sequences.

Recent developments in the theory of integrable models have provided the means of calculating dynamical correlation functions of some important observables in systems such as Heisenberg spin chains and one-dimensional atomic gases. This article explicitly describes how such calculations are generally implemented in the ABACUS C++ library, emphasizing the universality in treatment of different cases coming as a consequence of unifying features within the Bethe Ansatz.

The control of quantum phenomena is a topic that has carried out many challenging problems. Among others, the Hamiltonian identification, i.e, the inverse problem associated with the unknown features of a quantum system is still an open issue. In this work, we present an algorithm that enables to design a set of selective laser fields that can be used, in a second stage, to identify unknown parameters of quantum systems.

Based on speculations coming from statistical mechanics and the conjectured existence of critical states, I propose a simple heuristic in order to control the mutation probability and the population size of a genetic algorithm.

Recurrent relations for branching coefficients in affine Lie algebras integrable highest weight modules are studied. The decomposition algorithm based on the injection fan technique is developed for the case of an arbitrary reductive subalgebra. In particular we consider the situation where the Weyl denominator becomes singular with respect to the subalgebra. We demonstrate that for any reductive subalgebra it is possible to define the injection fan and the analogue of the Weyl numerator - the tools that describe explicitly the recurrent properties of branching coefficients. Possible applications of the fan technique in CFT models are considered.

The C-Algorithm introduced in [Chouikha2007] is designed to determine isochronous centers for Lienard-type differential systems, in the general real analytic case. However, it has a large complexity that prevents computations, even in the quartic polynomial case. The main result of this paper is an efficient algorithmic implementation of C-Algorithm, called ReCA (Reduced C-Algorithm). Moreover, an adapted version of it is proposed in the rational case. It is called RCA (Rational C-Algorithm) and is widely used in [BardetBoussaadaChouikhaStrelcyn2010] and [BoussaadaChouikhaStrelcyn2010] to find many new examples of isochronous centers for the Li\'enard type equation.

This paper presents an evolutionary algorithm with a new goal-sequence domination scheme for better decision support in multi-objective optimization. The approach allows the inclusion of advanced hard/soft priority and constraint information on each objective component, and is capable of incorporating multiple specifications with overlapping or non-overlapping objective functions via logical 'OR' and 'AND' connectives to drive the search towards multiple regions of trade-off. In addition, we propose a dynamic sharing scheme that is simple and adaptively estimated according to the on-line population distribution without needing any a priori parameter setting. Each feature in the proposed algorithm is examined to show its respective contribution, and the performance of the algorithm is compared with other evolutionary optimization methods. It is shown that the proposed algorithm has performed well in the diversity of evolutionary search and uniform distribution of non-dominated individuals along the final trade-offs, without significant computational effort. The algorithm is also applied to the design optimization of a practical servo control system for hard disk drives with a single voice-coil-motor actuator. Results of the evolutionary designed servo control system show a superior closed-loop performance compared to classical PID or RPT approaches.

To assist in the search for phenomenologically realistic models in the string landscape, we must develop tools for investigating all gauge charges, including U(1) charges, in string models. We introduce the process for constructing fermionic string models and present an algorithm for determining the U(1) gauge states and U(1) charges in weakly-coupled free fermionic heterotic string (WCFFHS) models given their matter and non-Abelian gauge content. We determine the computational complexity of this algorithm and discuss a particular implementation that can be used in conjunction with a framework developed at Baylor University for building WCFFHS models. We also present preliminary results regarding matter state uniqueness for a run of 1.4 million gauge models and find that U(1) charges affect the number of unique matter states in a significant minority of models constructed. We conclude by considering future avenues of investigation to which this algorithm may be applied.

We present the development and validation of a new multivariate $b$ jet identification algorithm ("$b$ tagger") used at the CDF experiment at the Fermilab Tevatron. At collider experiments, $b$ taggers allow one to distinguish particle jets containing $B$ hadrons from other jets. Employing feed-forward neural network architectures, this tagger is unique in its emphasis on using information from individual tracks. This tagger not only contains the usual advantages of a multivariate technique such as maximal use of information in a jet and tunable purity/efficiency operating points, but is also capable of evaluating jets with only a single track. To demonstrate the effectiveness of the tagger, we employ a novel method wherein we calculate the false tag rate and tag efficiency as a function of the placement of a lower threshold on a jet's neural network output value in $Z+1$ jet and $t\bar{t}$ candidate samples, rich in light flavor and $b$ jets, respectively.

Algorithm and code to produce sequences whose members obey Gaussian distribution function is reported. Discreet and limited number of groups are defined in the distribution function, where each group is represented only with one value instead of a range of value. The produced sequences are also checked back whether they still fit the discreet distribution function. Increasing of number of particles N increases the value of correlation coefficient R^2, but increasing number of groups M reduces it. Value R^2 = 1 can be found for N = 1000000 at least with M = 5000$ and for M = 10 at least with N = 1000.

We introduce PHEW (Parallel HiErarchical Watershed), a new segmentation algorithm to detect structures in astrophysical fluid simulations, and its implementation into the adaptive mesh refinement (AMR) code ramses. PHEW works on the density field defined on the adaptive mesh, and can thus be used on the gas density or the dark matter density after a projection of the particles onto the grid. The algorithm is based on a "watershed" segmentation of the computational volume into dense regions, followed by a merging of the segmented patches based on the saddle point topology of the density field. PHEW is capable of automatically detecting connected regions above the adopted density threshold, as well as the entire set of substructures within. Our algorithm is fully parallel and uses the MPI library. We describe in great detail the parallel algorithm and perform a scaling experiment which proves the capability of phew to run efficiently on massively parallel systems.

A method to determine the gamma correction curves for displays is presented. An empirical model is first constructed from exhaustive measurements of a few representative units. The model parameters for the remaining units are then fitted using only a few measurements. The method uses standard least-squares algorithms and is computationally light. Experimental results for a small sample of LCD displays are presented.

We propose a Viterbi-type trellis-search algorithm to implement the FSO photon-counting sequence receiver proposed in [1] more efficiently and a selective-store strategy to overcome the error floor problem observed therein.

In this paper, we present and analyze a simple and robust spectral algorithm for the stochastic block model with $k$ blocks, for any $k$ fixed. Our algorithm works with graphs having constant edge density, under an optimal condition on the gap between the density inside a block and the density between the blocks. As a co-product, we settle an open question posed by Abbe et. al. concerning censor block models.

The Baker-Campbell-Hausdorff series computes the quantity \begin{equation*} Z(X,Y)=\ln\left( e^X e^Y \right) = \sum_{n=1}^\infty z_n(X,Y), \end{equation*} where $X$ and $Y$ are not necessarily commuting, in terms of homogeneous multinomials $z_n(X,Y)$ of degree $n$. (This is essentially equivalent to computing the so-called Goldberg coefficients.) The Baker-Campbell-Hausdorff series is a general purpose tool of wide applicability in mathematical physics, quantum physics, and many other fields. The Reinsch algorithm for the truncated series permits one to calculate up to some fixed order $N$ by using $(N+1)\times(N+1)$ matrices. We show how to further simplify the Reinsch algorithm, making implementation (in principle) utterly straightforward. This helps provide a deeper understanding of the Goldberg coefficients and their properties. For instance we establish strict bounds (and some equalities) on the number of non-zero Goldberg coefficients. Unfortunately, we shall see that the number of terms in the multinomial $z_n(X,Y)$ often grows very rapidly (in fact exponentially) with the degree $n$. We also present some closely related results for the symmetric product \begin{equation*} S(X,Y)=\ln\left( e^{X/2} e^Y e^{X/2} \right) = \sum_{n=1}^\infty s_n(X,Y). \end{equation*} Variations on these themes are straightforward. For instance, one can just as easily consider the series \begin{equation*} L(X,Y)=\ln\left( e^{X} e^Y e^{-X} e^{-Y}\right) = \sum_{n=1}^\infty \ell_n(X,Y). \end{equation*} This type of series is of interest, for instance, when considering parallel transport around a closed curve. Several other related series are investigated.

We develop a forward-reverse EM (FREM) algorithm for estimating parameters that determine the dynamics of a discrete time Markov chain evolving through a certain measurable state space. As a key tool for the construction of the FREM method we develop forward-reverse representations for Markov chains conditioned on a certain terminal state. These representations may be considered as an extension of the earlier work Bayer and Schoenmakers [2013] on conditional diffusions. We proof almost sure convergence of our algorithm for a Markov chain model with curved exponential family structure. On the numerical side we give a complexity analysis of the forward-reverse algorithm by deriving its expected cost. Two application examples are discuss to demonstrate the scope of possible applications ranging from models based on continuous time processes to discrete time Markov chain models.

We describe and analyze a simple algorithm for principal component analysis and singular value decomposition, VR-PCA, which uses computationally cheap stochastic iterations, yet converges exponentially fast to the optimal solution. In contrast, existing algorithms suffer either from slow convergence, or computationally intensive iterations whose runtime scales with the data size. The algorithm builds on a recent variance-reduced stochastic gradient technique, which was previously analyzed for strongly convex optimization, whereas here we apply it to an inherently non-convex problem, using a very different analysis.

A distributed algorithm is described for solving a linear algebraic equation of the form $Ax=b$ assuming the equation has at least one solution. The equation is simultaneously solved by $m$ agents assuming each agent knows only a subset of the rows of the partitioned matrix $(A,b)$, the current estimates of the equation's solution generated by its neighbors, and nothing more. Each agent recursively updates its estimate by utilizing the current estimates generated by each of its neighbors. Neighbor relations are characterized by a time-dependent directed graph $\mathbb{N}(t)$ whose vertices correspond to agents and whose arcs depict neighbor relations. It is shown that for any matrix $A$ for which the equation has a solution and any sequence of "repeatedly jointly strongly connected graphs" $\mathbb{N}(t)$, $t=1,2,\ldots$, the algorithm causes all agents' estimates to converge exponentially fast to the same solution to $Ax=b$. It is also shown that the neighbor graph sequence must actually be repeatedly jointly strongly connected if exponential convergence is to be assured. A worst case convergence rate bound is derived for the case when $Ax=b$ has a unique solution. It is demonstrated that with minor modification, the algorithm can track the solution to $Ax = b$, even if $A$ and $b$ are changing with time, provided the rates of change of $A$ and $b$ are sufficiently small. It is also shown that in the absence of communication delays, exponential convergence to a solution occurs even if the times at which each agent updates its estimates are not synchronized with the update times of its neighbors. A modification of the algorithm is outlined which enables it to obtain a least squares solution to $Ax=b$ in a distributed manner, even if $Ax=b$ does not have a solution.

Accurate estimation of the position of network nodes is essential, e.g., in localization, geographic routing, and vehicular networks. Unfortunately, typical positioning techniques based on ranging or on velocity and angular measurements are inherently limited. To overcome the limitations of specific positioning techniques, the fusion of multiple and heterogeneous sensor information is an appealing strategy. In this paper, we investigate the fundamental performance of linear fusion of multiple measurements of the position of mobile nodes, and propose a new distributed recursive position estimator. The Cram\'er-Rao lower bounds for the parametric and a-posteriori cases are investigated. The proposed estimator combines information coming from ranging, speed, and angular measurements, which is jointly fused by a Pareto optimization problem where the mean and the variance of the localization error are simultaneously minimized. A distinguished feature of the method is that it assumes a very simple dynamical model of the mobility and therefore it is applicable to a large number of scenarios providing good performance. The main challenge is the characterization of the statistical information needed to model the Fisher information matrix and the Pareto optimization problem. The proposed analysis is validated by Monte Carlo simulations, and the performance is compared to several Kalman-based filters, commonly employed for localization and sensor fusion. Simulation results show that the proposed estimator outperforms the traditional approaches that are based on the extended Kalman filter when no assumption on the model of motion is used. In such a scenario, better performance is achieved by the proposed method, but at the price of an increased computational complexity.

The Laplace approximation calls for the computation of second derivatives at the likelihood maximum. When the maximum is found by the EM-algorithm, there is a convenient way to compute these derivatives. The likelihood gradient can be obtained from the EM-auxiliary, while the Hessian can be obtained from this gradient with the Pearlmutter trick.

With the success of dynamical mean field theories, solvers for quantum-impurity problems have become an important tool for the numerical study of strongly correlated systems. Continuous-time Quantum Monte Carlo sampling of the expansion in powers of the hybridization between the ``impurity'' and the bath provides a powerful solver when interactions are strong. Here we show that the usual updates that add or remove a pair of creation-annihilation operators are rigorously not ergodic for several classes of broken-symmetries that involve spatial components. We show that updates with larger numbers of simultaneous updates of pairs of creation-annihilation operators remedy this problem. As an example, we apply the four operator updates that are necessary for ergodicity to the case of d-wave superconductivity in plaquette dynamical mean-field theory for the one-band Hubbard model. While the results are qualitatively similar to those previously published, they are quantitatively better that previous ones, being closer to those obtained by other approaches.

Electing a leader is a classical problem in distributed computing system. Synchronization between processes often requires one process acting as a coordinator. If an elected leader node fails, the other nodes of the system need to elect another leader without much wasting of time. The bully algorithm is a classical approach for electing a leader in a synchronous distributed computing system, which is used to determine the process with highest priority number as the coordinator. In this paper, we have discussed the limitations of Bully algorithm and proposed a simple and efficient method for the Bully algorithm which reduces the number of messages during the election. Our analytical simulation shows that, our proposed algorithm is more efficient than the Bully algorithm with fewer messages passing and fewer stages.

We propose a new sufficient condition for verifying whether generic rank-r complex tensors of arbitrary order admit a unique decomposition as a linear combination of rank-1 tensors. A practical algorithm is proposed for verifying this condition, with which it was established that in all spaces of dimension less than 15000, with a few known exceptions, listed in the paper, generic identifiability holds for ranks up to one less than the generic rank of the space. This is the largest possible rank value for which generic identifiability can hold, except for spaces with a perfect shape. The algorithm can also verify the identifiability of a given specific rank-r decomposition, provided that it can be shown to correspond to a nonsingular point of the r-th order secant variety. For sufficiently small rank, which nevertheless improves upon the known bounds for specific identifiability, some local equations of this variety are known, allowing us to verify this property. As a particular example of our approach, we prove the identifiability of a specific 5x5x5 tensor of rank 7, which cannot be handled by the conditions recently provided in [I. Domanov and L. De Lathauwer, On the Uniqueness of the Canonical Polyadic Decomposition of third-order tensors--Part II: Uniqueness of the overall decomposition, SIAM J. Matrix Anal. Appl. 34(3), 2013]. Finally, we also present a surprising new class of weakly-defective Segre varieties that nevertheless turns out to admit a generically unique decomposition.

An additive 2-basis with range n is restricted if its largest element is n/2. Among the restricted 2-bases of given length k, the ones that have the greatest range are extremal restricted. We describe an algorithm that finds the extremal restricted 2-bases of a given length, and we list them for lengths up to k = 41.

We consider a mean-reverting stochastic volatility model which satisfies some relevant stylized facts of financial markets. We introduce an algorithm for the detection of peaks in the volatility profile, that we apply to the time series of Dow Jones Industrial Average and Financial Times Stock Exchange 100 in the period 1984-2013. Based on empirical results, we propose a bivariate version of the model, for which we find an explicit expression for the decay over time of cross-asset correlations between absolute returns. We compare our theoretical predictions with empirical estimates on the same financial time series, finding an excellent agreement.

Information spreads across social and technological networks, but often the network structures are hidden from us and we only observe the traces left by the diffusion processes, called cascades. Can we recover the hidden network structures from these observed cascades? What kind of cascades and how many cascades do we need? Are there some network structures which are more difficult than others to recover? Can we design efficient inference algorithms with provable guarantees? Despite the increasing availability of cascade data and methods for inferring networks from these data, a thorough theoretical understanding of the above questions remains largely unexplored in the literature. In this paper, we investigate the network structure inference problem for a general family of continuous-time diffusion models using an $l_1$-regularized likelihood maximization framework. We show that, as long as the cascade sampling process satisfies a natural incoherence condition, our framework can recover the correct network structure with high probability if we observe $O(d^3 \log N)$ cascades, where $d$ is the maximum number of parents of a node and $N$ is the total number of nodes. Moreover, we develop a simple and efficient soft-thresholding inference algorithm, which we use to illustrate the consequences of our theoretical results, and show that our framework outperforms other alternatives in practice.

An activity fundamental to science is building mathematical models. These models are used to both predict the results of future experiments and gain insight into the structure of the system under study. We present an algorithm that automates the model building process in a scientifically principled way. The algorithm can take observed trajectories from a wide variety of mechanical systems and, without any other prior knowledge or tuning of parameters, predict the future evolution of the system. It does this by applying the principle of least action and searching for the simplest Lagrangian that describes the system's behaviour. By generating this Lagrangian in a human interpretable form, it also provides insight into the working of the system.

Atmospheric refraction affects to various degrees exoplanet transit, lunar eclipse, as well as stellar occultation observations. Exoplanet retrieval algorithms often use analytical expressions for the column abundance along a ray traversing the atmosphere as well as for the deflection of that ray, which are first order approximations valid for low densities in a spherically symmetric homogeneous isothermal atmosphere. We derive new analytical formulae for both of these quantities, which are valid for higher densities, and use them to refine and validate a new ray tracing algorithm which can be used for arbitrary atmospheric temperature-pressure profiles. We illustrate with simple isothermal atmospheric profiles the consequences of our model for different planets: temperate Earth-like and Jovian-like planets, as well as HD189733b, and GJ1214b. We find that, for both hot exoplanets, our treatment of refraction does not make much of a difference to pressures as high as 10 atmosphere, but that it is important to consider the variation of gravity with altitude for GJ1214b. However, we find that the temperate atmospheres have an apparent scale height significantly smaller than their actual density scale height at densities larger than 1 amagat, thus increasing the difficulty of detecting spectral features originating in these regions. These denser atmospheric regions form a refractive boundary layer where column abundances and ray deflection increases dramatically with decreasing impact parameter. This refractive boundary layer mimics a surface, and none of the techniques mentioned above can probe atmospheric regions denser than about 4 amagat on these temperate planets.

Quantitative photoacoustic tomography (QPAT) is a recent hybrid imaging modality that couples optical tomography with ultrasound imaging to achieve high resolution imaging of optical properties of scattering media. Image reconstruction in QPAT is usually a two-step process. In the first step, the initial pressure field inside the medium, generated by the photoacoustic effect, is reconstructed using measured acoustic data. In the second step, this initial ultrasound pressure field datum is used to reconstruct optical properties of the medium. We propose in this work a one-step inversion algorithm for image reconstruction in QPAT that reconstructs the optical absorption coefficient directly from measured acoustic data. The algorithm can be used to recover simultaneously the absorption coefficient and the ultrasound speed of the medium from \emph{multiple} acoustic data sets, with appropriate \emph{a priori} bounds on the unknowns. We demonstrate, through numerical simulations based on synthetic data, the feasibility of the proposed reconstruction method.

We propose a bundle trust-region algorithm to minimize locally Lipschitz functions which are potentially nonsmooth and nonconvex. We prove global convergence of our method and show by way of an example that the classical convergence argument in trust-region methods based on the Cauchy point fails in the nonsmooth setting. Our method is tested experimentally on three problems in automatic control.

We present an effective heuristic for the Steiner Problem in Graphs. Its main elements are a multistart algorithm coupled with aggressive combination of elite solutions, both leveraging recently-proposed fast local searches. We also propose a fast implementation of a well-known dual ascent algorithm that not only makes our heuristics more robust (by quickly dealing with easier cases), but can also be used as a building block of an exact (branch-and-bound) algorithm that is quite effective for some inputs. On all graph classes we consider, our heuristic is competitive with (and sometimes more effective than) any previous approach with similar running times. It is also scalable: with long runs, we could improve or match the best published results for most open instances in the literature.

We prove that the approximation ratio of the greedy algorithm for the metric Traveling Salesman Problem is $\Theta(\log n)$. Moreover, we prove that the same result also holds for graphic, Euclidean, and rectilinear instances of the Traveling Salesman Problem. Finally we show that the approximation ratio of the Clarke-Wright savings heuristic for the metric Traveling Salesman Problem is $\Theta(\log n)$.

We propose a globally convergent alternating minimization (AM) algorithm for image reconstruction in transmission tomography, which extends automatic relevance determination (ARD) to Poisson noise models with Beer's law. The algorithm promotes solutions that are sparse in the pixel/voxel-differences domain by introducing additional latent variables, one for each pixel/voxel, and then learning these variables from the data using a hierarchical Bayesian model. Importantly, the proposed AM algorithm is free of any tuning parameters with image quality comparable to standard penalized likelihood methods. Our algorithm exploits optimization transfer principles which reduce the problem into parallel 1D optimization tasks (one for each pixel/voxel), making the algorithm feasible for large-scale problems. This approach considerably reduces the computational bottleneck of ARD associated with the posterior variances. Positivity constraints inherent in transmission tomography problems are also enforced. We demonstrate the performance of the proposed algorithm for x-ray computed tomography using synthetic and real-world datasets. The algorithm is shown to have much better performance than prior ARD algorithms based on approximate Gaussian noise models, even for high photon flux.

We consider the famous Strassen algorithm for fast multiplication of matrices. We show that this algorithm has a nontrivial finite group of automorphisms of order 36 (namely the direct product of two copies of the symmetric group on 3 symbols), or even 72, if we consider "extended" Strassen algorithm. This is an indirect evidence that the (unknown at present) optimal algorithm for multiplication of two size 3 by 3 matrices also may have a large automorphism group, and this may be a fruitful idea for a search of such an algorithm. In the beginning we give a brief introduction to the subject, to make the text accessible for specialists in the representation theory of finite groups.

In this paper, we are concerned with the weighted backup 2-center problem on a tree. The backup 2-center problem is a kind of center facility location problem, in which one is asked to deploy two facilities, with a given probability to fail, in a network. Given that the two facilities do not fail simultaneously, the goal is to find two locations, possibly on edges, that minimize the expected value of the maximum distance over all vertices to their closest functioning facility. In the weighted setting, each vertex in the network is associated with a nonnegative weight, and the distance from vertex $u$ to $v$ is weighted by the weight of $u$. With the strategy of prune-and-search, we propose a linear time algorithm, which is asymptotically optimal, to solve the weighted backup 2-center problem on a tree.

In Moving Horizon Estimation (MHE) the computed estimate is found by solving a constrained finite-time optimal estimation problem in real-time at each sample in a receding horizon fashion. The constrained estimation problem can be solved by, e.g., interior-point (IP) or active-set (AS) methods, where the main computational effort in both methods is known to be the computation of the search direction, i.e., the Newton step. This is often done using generic sparsity exploiting algorithms or serial Riccati recursions, but as parallel hardware is becoming more commonly available the need for parallel algorithms for computing the Newton step is increasing. In this paper a tailored, non-iterative parallel algorithm for computing the Newton step using the Riccati recursion is presented. The algorithm exploits the special structure of the Karush-Kuhn-Tucker system for the optimal estimation problem. As a result it is possible to obtain logarithmic complexity growth in the estimation horizon length, which can be used to reduce the computation time for IP and AS methods when applied to what is today considered as challenging estimation problems. Promising numerical results have been obtained using an ANSI-C implementation of the proposed algorithm running on true parallel hardware. Beyond MHE, due to similarities in the problem structure, the algorithm can be applied to various forms of on-line and off-line smoothing problems.

We develop coarse-graining tensor renormalization group algorithms to compute physical properties of two-dimensional lattice models on finite periodic lattices. Two different coarse-graining strategies, one based on the tensor renormalization group and the other based on the higher-order tensor renormalization group, are introduced. In order to optimize the tensor-network model globally, a sweeping scheme is proposed to account for the renormalization effect from the environment tensors under the framework of second renormalization group. We demonstrate the algorithms by the classical Ising model on the square lattice and the Kitaev model on the honeycomb lattice, and show that the finite-size algorithms achieve substantially more accurate results than the corresponding infinite-size ones.

This paper addresses the problem of separating audio sources from time-varying convolutive mixtures. We propose a probabilistic framework based on the local complex-Gaussian model combined with non-negative matrix factorization. The time-varying mixing filters are modeled by a continuous temporal stochastic process. We present a variational expectation-maximization (VEM) algorithm that employs a Kalman smoother to estimate the time-varying mixing matrix, and that jointly estimate the source parameters. The sound sources are then separated by Wiener filters constructed with the estimators provided by the VEM algorithm. Extensive experiments on simulated data show that the proposed method outperforms a block-wise version of a state-of-the-art baseline method.

In this work, we analyze the failing sets of the interval-passing algorithm (IPA) for compressed sensing. The IPA is an efficient iterative algorithm for reconstructing a k-sparse nonnegative n-dimensional real signal x from a small number of linear measurements y. In particular, we show that the IPA fails to recover x from y if and only if it fails to recover a corresponding binary vector of the same support, and also that only positions of nonzero values in the measurement matrix are of importance for success of recovery. Based on this observation, we introduce termatiko sets and show that the IPA fails to fully recover x if and only if the support of x contains a nonempty termatiko set, thus giving a complete (graph-theoretic) description of the failing sets of the IPA. Finally, we present an extensive numerical study showing that in many cases there exist termatiko sets of size strictly smaller than the stopping distance of the binary measurement matrix; even as low as half the stopping distance in some cases.

Political scientists often find themselves tracking amendments to political texts. As different actors weigh in, texts change as they are drafted and redrafted, reflecting political preferences and power. This study provides a novel solution to the prob- lem of detecting amendments to political text based upon minimum edit distances. We demonstrate the usefulness of two language-insensitive, transparent, and efficient minimum-edit-distance algorithms suited for the task. These algorithms are capable of providing an account of the types (insertions, deletions, substitutions, and trans- positions) and substantive amount of amendments made between version of texts. To illustrate the usefulness and efficiency of the approach we replicate two existing stud- ies from the field of legislative studies. Our results demonstrate that minimum edit distance methods can produce superior measures of text amendments to hand-coded efforts in a fraction of the time and resource costs.

Wavefront phase retrieval from a set of intensity measurements can be formulated as an optimization problem. Two nonconvex objective models (MLP and its variants LS) based on maximum likelihood estimation are investigated. We develop numerical optimization algorithms for real-valued function of complex variables and apply them to solve the wavefront phase retrieval problem efficiently. Numerical simulation is given with application to three wavefront phase retrieval problems. LS model shows better numerical performances than MLP model. An explanation for this is that the distribution of the eigenvalues of Hessian matrix of LS model is more clustered than MLP model. LBFGS shows more robust performance and takes fewer calculations than other line search methods.

We here introduce a novel classification approach adopted from the nonlinear model identification framework, which jointly addresses the feature selection and classifier design tasks. The classifier is constructed as a polynomial expansion of the original attributes and a model structure selection process is applied to find the relevant terms of the model. The selection method progressively refines a probability distribution defined on the model structure space, by extracting sample models from the current distribution and using the aggregate information obtained from the evaluation of the population of models to reinforce the probability of extracting the most important terms. To reduce the initial search space, distance correlation filtering can be applied as a preprocessing technique. The proposed method is evaluated and compared to other well-known feature selection and classification methods on standard benchmark classification problems. The results show the effectiveness of the proposed method with respect to competitor methods both in terms of classification accuracy and model complexity. The obtained models have a simple structure, easily amenable to interpretation and analysis.

The Long Term Evolution (LTE) as a mobile broadband technology supports a wide domain of communication services with different requirements. Therefore, scheduling of all flows from various applications in overload states in which the requested amount of bandwidth exceeds the limited available spectrum resources is a challenging issue. Accordingly, in this paper, a greedy algorithm is presented to evaluate user candidates which are waiting for scheduling and select an optimal set of the users to maximize system performance, without exceeding available bandwidth capacity. The greedy-knapsack algorithm is defined as an optimal solution to the resource allocation problem, formulated based on the fractional knapsack problem. A compromise between throughput and QoS provisioning is obtained by proposing a class-based ranking function, which is a combination of throughput and QoS related parameters defined for each application. The simulation results show that the proposed method provides high performance in terms of throughput, loss and delay for different classes of QoS over the existing ones, especially under overload traffic.

We propose a new algorithm for approximating the non-asymptotic second moment of the marginal likelihood estimate, or normalizing constant, provided by a particle filter. The computational cost of the new method is $O(M)$ per time step, independently of the number of particles $N$ in the particle filter, where $M$ is a parameter controlling the quality of the approximation. This is in contrast to $O(MN)$ for a simple averaging technique using $M$ i.i.d. replicates of a particle filter with $N$ particles. We establish that the approximation delivered by the new algorithm is unbiased, strongly consistent and, under standard regularity conditions, increasing $M$ linearly with time is sufficient to prevent growth of the relative variance of the approximation, whereas for the simple averaging technique it can be necessary to increase $M$ exponentially with time in order to achieve the same effect. Numerical examples illustrate performance in the context of a stochastic Lotka\textendash Volterra system and a simple AR(1) model.