Bayesian Structured Prediction Using Gaussian Processes

We introduce a conceptually novel structured prediction model, GPstruct, which is kernelized, non-parametric and Bayesian, by design. We motivate the model with respect to existing approaches, among others, conditional random fields (CRFs), maximum margin Markov networks (M3N), and structured support vector machines (SVMstruct), which embody only a subset of its properties. We present an inference procedure based on Markov Chain Monte Carlo. The framework can be instantiated for a wide range of structured objects such as linear chains, trees, grids, and other general graphs. As a proof of concept, the model is benchmarked on several natural language processing tasks and a video gesture segmentation task involving a linear chain structure. We show prediction accuracies for GPstruct which are comparable to or exceeding those of CRFs and SVMstruct.

Near field phase mapping exploiting intrinsic oscillations of aperture NSOM probe.

An innovative, simple, compact and low cost approach for phase mapping based on the intrinsic modulation of an aperture Near Field Scanning Optical Microscope probe is analyzed and experimentally demonstrated. Several nanoscale silicon waveguides are phase-mapped using this approach, and the different modes of propagation are obtained via Fourier analysis. The obtained measured results are in good agreement with the effective indexes of the modes calculated by electromagnetic simulations. Owing to its simplicity and effectiveness, the demonstrated system is a potential candidate for integration with current near field systems for the characterization of nanophotonic components and devices.

Near field phase mapping exploiting intrinsic oscillations of aperture NSOM probe

An innovative, simple, compact and low cost approach for phase mapping based on the intrinsic modulation of an aperture Near Field Scanning Optical Microscope probe is analyzed and experimentally demonstrated. Several nanoscale silicon waveguides are phase-mapped using this approach, and the different modes of propagation are obtained via Fourier analysis. The obtained measured results are in good agreement with the effective indexes of the modes calculated by electromagnetic simulations. Owing to its simplicity and effectiveness, the demonstrated system is a potential candidate for integration with current near field systems for the characterization of nanophotonic components and devices. © 2011 Optical Society of America.

Near field phase mapping exploiting intrinsic oscillations of NSOM probe

An innovative, simple compact and low cost approach for phase mapping based on the intrinsic modulation of a Near Field Scanning Optical Microscope probe is analyzed and experimentally demonstrated. © OSA/ CLEO 2011.

Nonparametric Bayesian Density Modeling with Gaussian Processes

We present the Gaussian process density sampler (GPDS), an exchangeable generative model for use in nonparametric Bayesian density estimation. Samples drawn from the GPDS are consistent with exact, independent samples from a distribution defined by a density that is a transformation of a function drawn from a Gaussian process prior. Our formulation allows us to infer an unknown density from data using Markov chain Monte Carlo, which gives samples from the posterior distribution over density functions and from the predictive distribution on data space. We describe two such MCMC methods. Both methods also allow inference of the hyperparameters of the Gaussian process.

Fixed-lag blind equalization and sequence estimation in digital communications systems using sequential importance sampling

We present methods for fixed-lag smoothing using Sequential Importance sampling (SIS) on a discrete non-linear, non-Gaussian state space system with unknown parameters. Our particular application is in the field of digital communication systems. Each input data point is taken from a finite set of symbols. We represent transmission media as a fixed filter with a finite impulse response (FIR), hence a discrete state-space system is formed. Conventional Markov chain Monte Carlo (MCMC) techniques such as the Gibbs sampler are unsuitable for this task because they can only perform processing on a batch of data. Data arrives sequentially, so it would seem sensible to process it in this way. In addition, many communication systems are interactive, so there is a maximum level of latency that can be tolerated before a symbol is decoded. We will demonstrate this method by simulation and compare its performance to existing techniques.

The ensemble statistics of the energy of a random system subjected to harmonic excitation

This paper is concerned with the ensemble statistics of the response to harmonic excitation of a single dynamic system such as a plate or an acoustic volume. Random point process theory is employed, and various statistical assumptions regarding the system natural frequencies are compared, namely: (i) Poisson natural frequency spacings, (ii) statistically independent Rayleigh natural frequency spacings, and (iii) natural frequency spacings conforming to the Gaussian orthogonal ensemble (GOE). The GOE is found to be the most realistic assumption, and simple formulae are derived for the variance of the energy of the system under either point loading or rain-on-the-roof excitation. The theoretical results are compared favourably with numerical simulations and experimental data for the case of a mass loaded plate. © 2003 Elsevier Ltd. All rights reserved.

Copula Processes

We define a copula process which describes the dependencies between arbitrarily many random variables independently of their marginal distributions. As an example, we develop a stochastic volatility model, Gaussian Copula Process Volatility (GCPV), to predict the latent standard deviations of a sequence of random variables. To make predictions we use Bayesian inference, with the Laplace approximation, and with Markov chain Monte Carlo as an alternative. We find both methods comparable. We also find our model can outperform GARCH on simulated and financial data. And unlike GARCH, GCPV can easily handle missing data, incorporate covariates other than time, and model a rich class of covariance structures.

Recurrent input transformations for Hidden Markov models

This paper presents a new architecture which integrates recurrent input transformations (RIT) and continuous density HMMs. The basic HMM structure is extended to accommodate recurrent neural networks which transform the input observations before they enter the Gaussian output distributions associated with the states of the HMM. During training the parameters of both HMM and RIT are simultaneously optimized according to the Maximum Mutual Information (MMI) criterion. Results are presented for the E-set recognition task which demonstrate the ability of recurrent input transformations to exploit longer term correlations in the speech signal and to give improved discrimination.

Blind equalization and sequence estimation using Markov chain Monte Carlo methods

This paper discusses the problem of restoring a digital input signal that has been degraded by an unknown FIR filter in noise, using the Gibbs sampler. A method for drawing a random sample of a sequence of bits is presented; this is shown to have faster convergence than a scheme by Chen and Li, which draws bits independently. ©1998 IEEE.

Random lasing in a bistable smectic A liquid crystal

In this work, we examine the phenomenon of random lasing from the smectic A liquid crystal phase. We summarise our results to date on random lasing from the smectic A phase including the ability to control the output from the sample using applied electric fields. In addition, diffuse random lasing is demonstrated from the electrohydrodynamic instabilities of a smectic A liquid crystal phase that has been doped with a low concentration of ionic impurities. Using a siloxane-based liquid crystal doped with ionic impurities and a laser dye, nonresonant random laser emission is observed from the highly scattering texture of the smectic A phase which is stable in zero-field. With the application of a low frequency alternating current electric field, turbulence is induced due to motion of the ions. This is accompanied by a decrease in the emission linewidth and an increase in the intensity of the laser emission. The benefit in this case is that a field is not required to maintain the texture as the scattering and homeotropic states are both stable in zero field. This offers a lower power consumption alternative to the electric-field induced static scattering sample.

Characterization of nano-composite M-2411/Y-123 thin films by electron backscatter diffraction and in-field critical current measurements

Thin films of nano-composite Y-Ba-Cu-O (YBCO) superconductors containing nano-sized, non-superconducting particles of Y2Ba 4CuMOx (M-2411 with M = Ag and Nb) have been prepared by the PLD technique. Electron backscatter diffraction (EBSD) has been used to analyze the crystallographic orientation of nano-particles embedded in the film microstructure. The superconducting YBa2Cu3O7 (Y-123) phase matrix is textured with a dominant (001) orientation for all samples, whereas the M-2411 phase exhibits a random orientation. Angular critical current measurements at various temperature (T) and applied magnetic field (B) have been performed on thin films containing different concentration of the M-2411 second phase. An increase in critical current density J c at T < 77 K and B < 6 T is observed for samples with low concentration of the second phase (2 mol % M-2411). Films containing 5 mol % Ag-2411 exhibit lower Jc than pure Y-123 thin films at all fields and temperatures. Samples with 5 mol % Nb-2411 show higher Jc(B) than phase pure Y-123 thin films for T < 77 K. © 2010 IOP Publishing Ltd.

Enhanced poisson sum representation for alpha-stable processes

In this paper we present Poisson sum series representations for α-stable (αS) random variables and a-stable processes, in particular concentrating on continuous-time autoregressive (CAR) models driven by α-stable Lévy processes. Our representations aim to provide a conditionally Gaussian framework, which will allow parameter estimation using Rao-Blackwellised versions of state of the art Bayesian computational methods such as particle filters and Markov chain Monte Carlo (MCMC). To overcome the issues due to truncation of the series, novel residual approximations are developed. Simulations demonstrate the potential of these Poisson sum representations for inference in otherwise intractable α-stable models. © 2011 IEEE.

Gaussian Process Regression Networks

We introduce a new regression framework, Gaussian process regression networks (GPRN), which combines the structural properties of Bayesian neural networks with the non-parametric flexibility of Gaussian processes. This model accommodates input dependent signal and noise correlations between multiple response variables, input dependent length-scales and amplitudes, and heavy-tailed predictive distributions. We derive both efficient Markov chain Monte Carlo and variational Bayes inference procedures for this model. We apply GPRN as a multiple output regression and multivariate volatility model, demonstrating substantially improved performance over eight popular multiple output (multi-task) Gaussian process models and three multivariate volatility models on benchmark datasets, including a 1000 dimensional gene expression dataset.

The analysis of stochastic EMC problems using the concept of diffuse field reciprocity

There has been much progress in recent years in the analysis of complex random vibro-acoustic systems, and general analysis methods have been developed which are based on the properties of diffuse wave fields. It is shown in the present paper that such methods can also be applied to high frequency EMC problems, avoiding the need for costly full wave solutions to Maxwell's equations in complex cavities. The theory behind the approach is outlined and then applied to the relatively simple case of a wiring system which is subject to reverberant electromagnetic wave excitation. © 2011 IEEE.