Search

In this study, we propose a non-linear random mapping model called GELM. The proposed model is based on a combination of the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model and the Extreme Learning Machine (ELM), and can be used to calculate Value-at-Risk (VaR). Alternatively, the GELM model is a non-parametric GARCH-type model. Compared with conventional models, such as the GARCH models, ELM, and Support Vector Machine (SVM), the computational results confirm that the GELM model performs better in volatility forecasting and VaR calculation in terms of efficiency and accuracy. Thus, the GELM model can be an essential tool for risk management and stress testing.

Similar text algorithm is the main method for news topic extraction and online public opinion hot spot tracking. The traditional text similarity method has the problems of high data dimension and high computational complexity. This paper proposes a method of news text clustering based on hash feature and random mapping. The method based on hash feature reduces the dimensionality of the text. At the same time, the article clustering preprocessing based on random mapping can speed up retrieval. Experiments show that the algorithm proposed in this paper performs well on related data sets.

Recurrent neural networks (RNNs) remain challenging, and there is still a lack of long-term memory or learning ability in sequential data classification and prediction. In this paper, we propose a flexible recurrent model, BIdirectional COnvolutional RaNdom RNNs (BICORN-RNNs), incorporating a series of sub-models: random projection, convolutional operation, and bidirectional transmission. These subcategories advance classification accuracy, which was limited by the gradient vanishing and the exploding problem. Experiments on public time series datasets demonstrate that our proposed method substantially outperforms a variety of existing models. Furthermore, the coordination of the accuracy and efficiency concerning a variety of factors, including SNR, length, data missing, and overlapping, is also discussed.

Channel polarisation results are extended to the case of communications over parallel channels, where the channel state information is known to both the encoder and decoder. Given a set of parallel binary-input discrete memoryless channels (B-DMCs), by performing the channel polarising transformation over independent copies of these component channels, we obtain a second set of synthesised binary-input channels. Similar to the single-channel case, we prove that as the size of the transformation goes infinity, some of the resulting channels tend to completely noised, and the others tend to noise-free, where the fraction of the latter approaches the average symmetric capacity of the underlying component channels. For finite-length polar coding over parallel channels, performance is found to be relied heavily on the specific channelmapping scheme. To avoid exhaustive searching, an empirically good scheme that is called equal-capacity partition channel mapping is proposed and numerical results show that the proposed scheme significantly outperforms random mapping. Further, utilising the above results, a polar coding method for arbitrary code length is proposed, which has potential applications in practical systems.

Version éditeur de cette publication à l'adresse suivante : http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6514655 To solve stochastic problems with geometric uncertainties, one can transform the original problem in a domain with stochastic boundaries and interfaces to a problem defined in a deterministic domain with uncertainties in the material behavior. The latter problem is then discretized. There exist infinitely many random mappings that lead to identical results in the continuous domain but not in the discretized domain. In this paper, an a priori error indicator is proposed for electromagnetic problems with scalar and vector potential formulations. This leads to criteria for selecting random mappings that reduce the numerical error. In an illustrative numerical example, the proposed a priori error indicator is compared with an a posteriori estimator for both potential formulations This work is supported by the program MEDEE funded by the Nord Pas de Calais council and the European Community and supported in part by the National Science Foundation under Grant No. 1216927.

Let $T_{n,m}=\mathbb Z_n\times\mathbb Z_m$, and define a random mapping $\phi\colon T_{n,m}\to T_{n,m}$ by $\phi(x,y)=(x+1,y)$ or $(x,y+1)$ independently over $x$ and $y$ and with equal probability. We study the orbit structure of such ``quenched random walks'' $\phi$ in the limit $m,n\to\infty$, and show how it depends sensitively on the ratio $m/n$. For $m/n$ near a rational $p/q$, we show that there are likely to be on the order of $\sqrt{n}$ cycles, each of length O(n), whereas for $m/n$ far from any rational with small denominator, there are a bounded number of cycles, and for typical $m/n$ each cycle has length on the order of $n^{4/3}$., Comment: 22 pages, 3 figures

We consider random walks on the mapping class group whose support generates a non-elementary subgroup and contains a pseudo-Anosov map whose invariant Teichm\"uller geodesic is in the principal stratum. For such random walks, we show that mapping classes along almost every infinite sample path are eventually pseudo-Anosov, with invariant Teichm\"uller geodesics in the principal stratum. This provides an answer to a question of Kapovich and Pfaff., Comment: 12 pages, 2 figures

We investigate the problem of the predictability of random variable $Y$ under a privacy constraint dictated by random variable $X$, correlated with $Y$, where both predictability and privacy are assessed in terms of the minimum mean-squared error (MMSE). Given that $X$ and $Y$ are connected via a binary-input symmetric-output (BISO) channel, we derive the \emph{optimal} random mapping $P_{Z|Y}$ such that the MMSE of $Y$ given $Z$ is minimized while the MMSE of $X$ given $Z$ is greater than $(1-\epsilon)\mathsf{var}(X)$ for a given $\epsilon\geq 0$. We also consider the case where $(X,Y)$ are continuous and $P_{Z|Y}$ is restricted to be an additive noise channel., Comment: 9 pages, 3 figures

A random mapping (T n ;q) of a finite set V, V = {1,2,…,n}, into itself assigns independently to each i ∊ V its unique image j ∊ V with probability q if i = j and with probability P = (1-q)/(n−1) if i ≠ j. Three versions of epidemic processes on a random digraph G T representing (T n ;q) are studied. The exact probability distributions of the total number of infected elements as well as the threshold functions for these epidemic processes are determined.

We consider a random walk on the mapping class group of a surface of finite type. We assume that the random walk is determined by a probability measure whose support is finite and generates a non-elementary subgroup $H$. We further assume that $H$ is not consisting only of lifts with respect to any one covering. Then we prove that the probability that such a random walk gives a non-minimal mapping class in its fibered commensurability class decays exponentially. As an application of the minimality, we prove that for the case where a surface has at least one puncture, the probability that a random walk gives mapping classes with arithmetic mapping tori decays exponentially. We also prove that a random walk gives rise to asymmetric mapping tori with exponentially high probability for closed case., Comment: Earlier versions contain a gap in Section 4 (Lemma 4.8 of v1 and v2 is false). 25 pages (10 pages longer)

In this paper we characterise the distributions of the number of predecessors and of the number of successors of a given set of vertices, A, in the random mapping model, T n D̂ (see Hansen and Jaworski (2008)), with exchangeable in-degree sequence (D̂ 1,D̂ 2,…,D̂ n ). We show that the exact formulae for these distributions and their expected values can be given in terms of the distributions of simple functions of the in-degree variables D̂ 1,D̂ 2,…,D̂ n . As an application of these results, we consider two special examples of T n D̂ which correspond to random mappings with preferential and anti-preferential attachment, and determine the exact distributions for the number of predecessors and the number of successors in these cases. We also characterise, for these two special examples, the asymptotic behaviour of the expected numbers of predecessors and successors and interpret these results in terms of the threshold behaviour of epidemic processes on random mapping graphs. The families of discrete distributions obtained in this paper are also of independent interest.

The numerical solution of partial differential equations onto random domains can be done by using a mapping transforming this random domain into a deterministic domain. The issue is then to determine this one to one random mapping. In this paper, we present two methods-one based on the resolution of the Laplace equations, one based on a geometric transformation-to determine the random mapping. A stochastic magnetostatic example is treated to compare these methods. This work is supported by the program MEDEE funded by the Nord Pas de Calais council and the European Community

A simple identity for the incomplete factorial of sums of zero-one variables is exploited to provide the factorial moments of the number of components and the number of cyclical elements of the random mapping (T, {pi }) considered by Ross (1981).

International audience; In this paper, we study the problem of optimizing the throughput for micro-factories subject to failures. The challenge consists in mapping several tasks onto a set of machines. The originality of our approach is the failure model for such applications in which tasks are subject to failures rather than machines. If there is exactly one task per machine in the mapping, then we prove that the optimal solution can be computed in polynomial time. However, the problem becomes NP-hard if several tasks can be assigned to the same machine. Several polynomial time heuristics are presented for the most realistic {\em specialized} setting, in which tasks of a same type can be mapped onto the same machine. Experimental results show that the best heuristics obtain a good throughput, much better than the throughput obtained with a random mapping. Moreover, we obtain a throughput close to the optimal solution in the particular cases on which the optimal throughput can be computed.

A random mapping partitions the set {1, 2, ···, m} into components, where i and j are in the same component if some functional iterate of i equals some functional iterate of j. We consider various functionals of these partitions and of samples from it, including the number of components of ‘small' size and of size O(m) as m → ∞the size of the largest component, the number of components, and various symmetric functionals of the normalized component sizes. In many cases exact results, while available, are uniformative, and we consider various approximations. Numerical and simulation results are also presented. A central tool for many calculations is the ‘frequency spectrum', both exact and asymptotic.

An acyclic mapping from an $n$ element set into itself is a mapping $\phi$ such that if $\phi^k(x) = x$ for some $k$ and $x$, then $\phi(x) = x$. Equivalently, $\phi^\ell = \phi^{\ell+1} = ...$ for $\ell$ sufficiently large. We investigate the behavior as $n \to \infty$ of a Markov chain on the collection of such mappings. At each step of the chain, a point in the $n$ element set is chosen uniformly at random and the current mapping is modified by replacing the current image of that point by a new one chosen independently and uniformly at random, conditional on the resulting mapping being again acyclic. We can represent an acyclic mapping as a directed graph (such a graph will be a collection of rooted trees) and think of these directed graphs as metric spaces with some extra structure. Heuristic calculations indicate that the metric space valued process associated with the Markov chain should, after an appropriate time and ``space'' rescaling, converge as $n \to \infty$ to a real tree ($\R$-tree) valued Markov process that is reversible with respect to a measure induced naturally by the standard reflected Brownian bridge. The limit process, which we construct using Dirichlet form methods, is a Hunt process with respect to a suitable Gromov-Hausdorff-like metric. This process is similar to one that appears in earlier work by Evans and Winter as the limit of chains involving the subtree prune and regraft tree (SPR) rearrangements from phylogenetics., Comment: 26 pages, 4 figures

We study the asymptotics of the $p$-mapping model of random mappings on $[n]$ as $n$ gets large, under a large class of asymptotic regimes for the underlying distribution $p$. We encode these random mappings in random walks which are shown to converge to a functional of the exploration process of inhomogeneous random trees, this exploration process being derived (Aldous-Miermont-Pitman 2003) from a bridge with exchangeable increments. Our setting generalizes previous results by allowing a finite number of ``attracting points'' to emerge., 16 pages

The Joyal bijection between doubly-rooted trees and mappings can be lifted to a transformation on function space which takes tree-walks to mapping-walks. Applying known results on weak convergence of random tree walks to Brownian excursion, we give a conceptually simpler rederivation of the Aldous-Pitman (1994) result on convergence of uniform random mapping walks to reflecting Brownian bridge, and extend this result to random p-mappings. © 2004 Applied Probability Trust.

A random mapping of a countable set N into itself is studied. An explicit formula is found for the probability that a subset of N will not contain cycles of the mapping. This formula is applied then for a detailed asymptotic analysis of two special mappings.

A random mapping (T,Pj) of a finite set V into itself is studied. We give a new proof of the fundamental lemma of [6]. Our method leads to the derivation of several results which cannot be deduced from [6]. In particular we determine the distribution of the number of components, cyclical points and ancestors of a given point.

(1954). The Expected Number of Components Under a Random Mapping Function. The American Mathematical Monthly: Vol. 61, No. 6, pp. 392-397.

Mapping patterns may be represented by unlabelled directed graphs in which each point has outdegree one. We consider the uniform probability measure on the set of all mapping patterns on n points and derive the limiting distribution of the size of the largest tree as n → ∞. It is also shown that the asymptotic results concerning the structure of trees of a random mapping pattern coincide with those obtained earlier for the graphs of mappings on labelled point-sets.

We use Watson's lemma to obtain an asymptotic expansion for the probability that a random mapping is connected.