Model selection approaches for non-linear system identification: a review

The identification of non-linear systems using only observed finite datasets has become a mature research area over the last two decades. A class of linear-in-the-parameter models with universal approximation capabilities have been intensively studied and widely used due to the availability of many linear-learning algorithms and their inherent convergence conditions. This article presents a systematic overview of basic research on model selection approaches for linear-in-the-parameter models. One of the fundamental problems in non-linear system identification is to find the minimal model with the best model generalisation performance from observational data only. The important concepts in achieving good model generalisation used in various non-linear system-identification algorithms are first reviewed, including Bayesian parameter regularisation and models selective criteria based on the cross validation and experimental design. A significant advance in machine learning has been the development of the support vector machine as a means for identifying kernel models based on the structural risk minimisation principle. The developments on the convex optimisation-based model construction algorithms including the support vector regression algorithms are outlined. Input selection algorithms and on-line system identification algorithms are also included in this review. Finally, some industrial applications of non-linear models are discussed.

Do All Flares Have White-Light Emission?

High-cadence, multiwavelength optical observations of a solar active region (NOAA AR 10969), obtained with the Swedish Solar Telescope, are presented. Difference imaging of white light continuum data reveals a white-light brightening, 2 minutes in duration, linked to a cotemporal and cospatial C2.0 flare event. The flare kernel observed in the white-light images has a diameter of 300 km, thus rendering it below the resolution limit of most space-based telescopes. Continuum emission is present only during the impulsive stage of the flare, with the effects of chromospheric emission subsequently delayed by approximate to 2 minutes. The localized flare emission peaks at 300% above the quiescent flux. This large, yet tightly confined, increase in emission is only resolvable due to the high spatial resolution of the Swedish Solar Telescope. An investigation of the line-of-sight magnetic field derived from simultaneous MDI data shows that the continuum brightening is located very close to a magnetic polarity inversion line. In addition, an Ha flare ribbon is directed along a region of rapid magnetic energy change, with the footpoints of the ribbon remaining cospatial with the observed white-light brightening throughout the duration of the flare. The observed flare parameters are compared with current observations and theoretical models for M- and X-class events and we determine the observed white-light emission is caused by radiative back-warming. We suggest that the creation of white-light emission is a common feature of all solar flares.

La visión artificial, un nuevo aliado para el análisis de imágenes artísticas

En este estudio se evalúa el rendimiento de los métodos de Bag-of-Visualterms (BOV) para la clasificación automática de imágenes digitales de la base de datos del artista Miquel Planas. Estas imágenes intervienen en la ideación y diseño de su producción escultórica. Constituye un interesante desafío dada la dificultad de la categorización de escenas cuando éstas difieren más por los contenidos semánticos que por los objetos que contienen. Hemos empleado un método de reconocimiento basado en Kernels introducido por Lazebnik, Schmid y Ponce en 2006. Los resultados son prometedores, en promedio, la puntuación del rendimiento es aproximadamente del 70%. Los experimentos sugieren que la categorización automática de imágenes basada en métodos de visión artificial puede proporcionar principios objetivos en la catalogación de imágenes y que los resultados obtenidos pueden ser aplicados en diferentes campos de la creación artística.

Analysis of information leakage from encrypted Skype conversations

Voice over IP (VoIP) has experienced a tremendous growth over the last few years and is now widely used among the population and for business purposes. The security of such VoIP systems is often assumed, creating a false sense of privacy. This paper investigates in detail the leakage of information from Skype, a widely used and protected VoIP application. Experiments have shown that isolated phonemes can be classified and given sentences identified. By using the dynamic time warping (DTW) algorithm, frequently used in speech processing, an accuracy of 60% can be reached. The results can be further improved by choosing specific training data and reach an accuracy of 83% under specific conditions. The initial results being speaker dependent, an approach involving the Kalman filter is proposed to extract the kernel of all training signals.

Improved Nonlinear PCA for Process Monitoring Using Support Vector Data Description

Nonlinear principal component analysis (PCA) based on neural networks has drawn significant attention as a monitoring tool for complex nonlinear processes, but there remains a difficulty with determining the optimal network topology. This paper exploits the advantages of the Fast Recursive Algorithm, where the number of nodes, the location of centres, and the weights between the hidden layer and the output layer can be identified simultaneously for the radial basis function (RBF) networks. The topology problem for the nonlinear PCA based on neural networks can thus be solved. Another problem with nonlinear PCA is that the derived nonlinear scores may not be statistically independent or follow a simple parametric distribution. This hinders its applications in process monitoring since the simplicity of applying predetermined probability distribution functions is lost. This paper proposes the use of a support vector data description and shows that transforming the nonlinear principal components into a feature space allows a simple statistical inference. Results from both simulated and industrial data confirm the efficacy of the proposed method for solving nonlinear principal component problems, compared with linear PCA and kernel PCA.

Electromagnetic wave diffraction from an infinite array of complex-shaped microstrip reflectors

A numerical-analytical method is developed for solving surface integral equations (IEs) describing electromagnetic wave diffraction from arrays of complex-shaped planar reflectors. Solutions to these equations are regularized via analytical transformation of the separated singular part of the matrix kernel. Basis functions satisfying the metal-edge condition are determined on the entire surface of the complex region. The amplitude and phase responses of arrays consisting of polygonal reflectors are numerically investigated.

The Geography of Conflict and Death in Belfast, Northern Ireland

The conflict known as the oTroubleso in Northern Ireland began during the late 1960s and is defined by political and ethno-sectarian violence between state, pro-state, and anti-state forces. Reasons for the conflict are contested and complicated by social, religious, political, and cultural disputes, with much of the debate concerning the victims of violence hardened by competing propaganda-conditioning perspectives. This article introduces a database holding information on the location of individual fatalities connected with the contemporary Irish conflict. For each victim, it includes a demographic profile, home address, manner of death, and the organization responsible. Employing geographic information system (GIS) techniques, the database is used to measure, map, and analyze the spatial distribution of conflict-related deaths between 1966 and 2007 across Belfast, the capital city of Northern Ireland, with respect to levels of segregation, social and economic deprivation, and interfacing. The GIS analysis includes a kernel density estimator designed to generate smooth intensity surfaces of the conflict-related deaths by both incident and home locations. Neighborhoods with high-intensity surfaces of deaths were those with the highest levels of segregation ( 90 percent Catholic or Protestant) and deprivation, and they were located near physical barriers, the so-called peacelines, between predominantly Catholic and predominantly Protestant communities. Finally, despite the onset of peace and the formation of a power-sharing and devolved administration (the Northern Ireland Assembly), disagreements remain over the responsibility and ocommemorationo of victims, sentiments that still uphold division and atavistic attitudes between spatially divided Catholic and Protestant populations.

Model Selection in SVMs Using Differential Evolution

To improve the performance of classification using Support Vector Machines (SVMs) while reducing the model selection time, this paper introduces Differential Evolution, a heuristic method for model selection in two-class SVMs with a RBF kernel. The model selection method and related tuning algorithm are both presented. Experimental results from application to a selection of benchmark datasets for SVMs show that this method can produce an optimized classification in less time and with higher accuracy than a classical grid search. Comparison with a Particle Swarm Optimization (PSO) based alternative is also included.

Chromospheric velocities of a C-class flare

Aims. We use high spatial and temporal resolution observations from the Swedish Solar Telescope to study the chromospheric velocities of a C-class flare originating from active region NOAA 10969.
Methods. A time-distance analysis is employed to estimate directional velocity components in Hα and Ca ii K image sequences. Also, imaging spectroscopy has allowed us to determine flare-induced line-of-sight velocities. A wavelet analysis is used to analyse the periodic nature of associated flare bursts.
Results. Time-distance analysis reveals velocities as high as 64 km s-1 along the flare ribbon and 15 km s-1 perpendicular to it. The velocities are very similar in both the Hα and Ca ii K time series. Line-of-sight Hα velocities are red-shifted with values up to 17 km s-1. The high spatial and temporal resolution of the observations have allowed us to detect velocities significantly higher than those found in earlier studies. Flare bursts with a periodicity of ≈60 s are also detected. These bursts are similar to the quasi-periodic oscillations observed at hard X-ray and radio wavelength data.
Conclusions. Some of the highest velocities detected in the solar atmosphere are presented. Line-of-sight velocity maps show considerable mixing of both the magnitude and direction of velocities along the flare path. A change in direction of the velocities at the flare kernel has also been detected which may be a signature of chromospheric evaporation.

Three-dimensional delayed-detonation models with nucleosynthesis for type ia supernovae

We present results for a suite of 14 three-dimensional, high-resolution hydrodynamical simulations of delayed-detonation models of Type Ia supernova (SN Ia) explosions. This model suite comprises the first set of three-dimensional SN Ia simulations with detailed isotopic yield information. As such, it may serve as a data base for Chandrasekhar-mass delayed-detonation model nucleosynthetic yields and for deriving synthetic observables such as spectra and light curves. We employ aphysically motivated, stochastic model based on turbulent velocity fluctuations and fuel density to calculate in situ the deflagration-to-detonation transition probabilities. To obtain different strengths of the deflagration phase and thereby different degrees of pre-expansion, we have chosen a sequence of initial models with 1, 3, 5, 10, 20, 40, 100, 150, 200, 300 and 1600 (two different realizations) ignition kernels in a hydrostatic white dwarf with a central density of 2.9 × 10 g cm, as well as one high central density (5.5 × 10 g cm) and one low central density (1.0 × 10 g cm) rendition of the 100 ignition kernel configuration. For each simulation, we determined detailed nucleosynthetic yields by postprocessing10 tracer particles with a 384 nuclide reaction network. All delayed-detonation models result in explosions unbinding thewhite dwarf, producing a range of 56Ni masses from 0.32 to 1.11M. As a general trend, the models predict that the stableneutron-rich iron-group isotopes are not found at the lowest velocities, but rather at intermediate velocities (~3000×10 000 km s) in a shell surrounding a Ni-rich core. The models further predict relatively low-velocity oxygen and carbon, with typical minimum velocities around 4000 and 10 000 km s, respectively. © 2012 The Authors. Published by Oxford University Press on behalf of the Royal Astronomical Society.

K-Theory of Azumaya Algebras over Schemes

Let X be a connected, noetherian scheme and A{script} be a sheaf of Azumaya algebras on X, which is a locally free O{script}-module of rank a. We show that the kernel and cokernel of K(X) ? K(A{script}) are torsion groups with exponent a for some m and any i = 0, when X is regular or X is of dimension d with an ample sheaf (in this case m = d + 1). As a consequence, K(X, Z/m) ? K(A{script}, Z/m), for any m relatively prime to a. © 2013 Copyright Taylor and Francis Group, LLC.

Hexavalent Chromium Adsorption by a Novel Activated Carbon Prepared by Microwave Activation

Microwave heating reduces the preparation time and improves the adsorption quality of activated carbon. In this study, activated carbon was prepared by impregnation of palm kernel fiber with phosphoric acid followed by microwave activation. Three different types of activated carbon were prepared, having high surface areas of 872 m2 g-1, 1256 m2 g-1, and 952 m2 g-1 and pore volumes of 0.598 cc g-1, 1.010 cc g-1, and 0.778 cc g-1, respectively. The combined effects of the different process parameters, such as the initial adsorbate concentration, pH, and temperature, on adsorption efficiency were explored with the help of Box-Behnken design for response surface methodology (RSM). The adsorption rate could be expressed by a polynomial equation as the function of the independent variables. The hexavalent chromium adsorption rate was found to be 19.1 mg g-1 at the optimized conditions of the process parameters, i.e., initial concentration of 60 mg L-1, pH of 3, and operating temperature of 50 oC. Adsorption of Cr(VI) by the prepared activated carbon was spontaneous and followed second-order kinetics. The adsorption mechanism can be described by the Freundlich Isotherm model. The prepared activated carbon has demonstrated comparable performance to other available activated carbons for the adsorption of Cr(VI).

Rigorous Specification and Low-Latency Implementation of Technical Market Indicators

Technical market indicators are tools used by technical an- alysts to understand trends in trading markets. Technical (market) indicators are often calculated in real-time, as trading progresses. This paper presents a mathematically- founded framework for calculating technical indicators. Our framework consists of a domain specific language for the un- ambiguous specification of technical indicators, and a run- time system based on Click, for computing the indicators. We argue that our solution enhances the ease of program- ming due to aligning our domain-specific language to the mathematical description of technical indicators, and that it enables executing programs in kernel space for decreased latency, without exposing the system to users’ programming errors.

Learning from time series:Supervised Aggregative Feature Extraction

Many modeling problems require to estimate a scalar output from one or more time series. Such problems are usually tackled by extracting a fixed number of features from the time series (like their statistical moments), with a consequent loss in information that leads to suboptimal predictive models. Moreover, feature extraction techniques usually make assumptions that are not met by real world settings (e.g. uniformly sampled time series of constant length), and fail to deliver a thorough methodology to deal with noisy data. In this paper a methodology based on functional learning is proposed to overcome the aforementioned problems; the proposed Supervised Aggregative Feature Extraction (SAFE) approach allows to derive continuous, smooth estimates of time series data (yielding aggregate local information), while simultaneously estimating a continuous shape function yielding optimal predictions. The SAFE paradigm enjoys several properties like closed form solution, incorporation of first and second order derivative information into the regressor matrix, interpretability of the generated functional predictor and the possibility to exploit Reproducing Kernel Hilbert Spaces setting to yield nonlinear predictive models. Simulation studies are provided to highlight the strengths of the new methodology w.r.t. standard unsupervised feature selection approaches. © 2012 IEEE.

Efficient Marginal Likelihood Computation for Gaussian Process Regression

In a Bayesian learning setting, the posterior distribution of a predictive model arises from a trade-off between its prior distribution and the conditional likelihood of observed data. Such distribution functions usually rely on additional hyperparameters which need to be tuned in order to achieve optimum predictive performance; this operation can be efficiently performed in an Empirical Bayes fashion by maximizing the posterior marginal likelihood of the observed data. Since the score function of this optimization problem is in general characterized by the presence of local optima, it is necessary to resort to global optimization strategies, which require a large number of function evaluations. Given that the evaluation is usually computationally intensive and badly scaled with respect to the dataset size, the maximum number of observations that can be treated simultaneously is quite limited. In this paper, we consider the case of hyperparameter tuning in Gaussian process regression. A straightforward implementation of the posterior log-likelihood for this model requires O(N^3) operations for every iteration of the optimization procedure, where N is the number of examples in the input dataset. We derive a novel set of identities that allow, after an initial overhead of O(N^3), the evaluation of the score function, as well as the Jacobian and Hessian matrices, in O(N) operations. We prove how the proposed identities, that follow from the eigendecomposition of the kernel matrix, yield a reduction of several orders of magnitude in the computation time for the hyperparameter optimization problem. Notably, the proposed solution provides computational advantages even with respect to state of the art approximations that rely on sparse kernel matrices.