Calibrating the Gaussian multi-target tracking model

We present novel batch and online (sequential) versions of the expectation-maximisation (EM) algorithm for inferring the static parameters of a multiple target tracking (MTT) model. Online EM is of particular interest as it is a more practical method for long data sets since in batch EM, or a full Bayesian approach, a complete browse of the data is required between successive parameter updates. Online EM is also suited to MTT applications that demand real-time processing of the data. Performance is assessed in numerical examples using simulated data for various scenarios. For batch estimation our method significantly outperforms an existing gradient based maximum likelihood technique, which we show to be significantly biased. © 2014 Springer Science+Business Media New York.

Symmetrical 2-D hermite-gaussian square launch for high bit rate transmission in multimode fiber links

A 2-D Hermite-Gaussian square launch is demonstrated to show improved systems capacity over multimode fiber links. It shows a bandwidth improvement over both center and offset launches and exhibits ±5 ìm misalignment tolerance. © OSA/OFC/NFOEC 2011.

Avoiding pathologies in very deep networks

Choosing appropriate architectures and regularization strategies of deep networks is crucial to good predictive performance. To shed light on this problem, we analyze the analogous problem of constructing useful priors on compositions of functions. Specifically, we study the deep Gaussian process, a type of infinitely-wide, deep neural network. We show that in standard architectures, the representational capacity of the network tends to capture fewer degrees of freedom as the number of layers increases, retaining only a single degree of freedom in the limit. We propose an alternate network architecture which does not suffer from this pathology. We also examine deep covariance functions, obtained by composing infinitely many feature transforms. Lastly, we characterize the class of models obtained by performing dropout on Gaussian processes.

Fault tolerant control using Gaussian processes and model predictive control

Essential ingredients for fault-tolerant control are the ability to represent system behaviour following the occurrence of a fault, and the ability to exploit this representation for deciding control actions. Gaussian processes seem to be very promising candidates for the first of these, and model predictive control has a proven capability for the second. We therefore propose to use the two together to obtain fault-tolerant control functionality. Our proposal is illustrated by several reasonably realistic examples drawn from flight control. © 2013 IEEE.

First principles interatomic potential for tungsten based on Gaussian process regression

An accurate description of atomic interactions, such as that provided by first principles quantum mechanics, is fundamental to realistic prediction of the properties that govern plasticity, fracture or crack propagation in metals. However, the computational complexity associated with modern schemes explicitly based on quantum mechanics limits their applications to systems of a few hundreds of atoms at most. This thesis investigates the application of the Gaussian Approximation Potential (GAP) scheme to atomistic modelling of tungsten - a bcc transition metal which exhibits a brittle-to-ductile transition and whose plasticity behaviour is controlled by the properties of $\frac{1}{2} \langle 111 \rangle$ screw dislocations. We apply Gaussian process regression to interpolate the quantum-mechanical (QM) potential energy surface from a set of points in atomic configuration space. Our training data is based on QM information that is computed directly using density functional theory (DFT). To perform the fitting, we represent atomic environments using a set of rotationally, permutationally and reflection invariant parameters which act as the independent variables in our equations of non-parametric, non-linear regression. We develop a protocol for generating GAP models capable of describing lattice defects in metals by building a series of interatomic potentials for tungsten. We then demonstrate that a GAP potential based on a Smooth Overlap of Atomic Positions (SOAP) covariance function provides a description of the $\frac{1}{2} \langle 111 \rangle$ screw dislocation that is in agreement with the DFT model. We use this potential to simulate the mobility of $\frac{1}{2} \langle 111 \rangle$ screw dislocations by computing the Peierls barrier and model dislocation-vacancy interactions to QM accuracy in a system containing more than 100,000 atoms.

Gaussian processes for POMDP-based dialogue manager optimization

A partially observable Markov decision process (POMDP) has been proposed as a dialog model that enables automatic optimization of the dialog policy and provides robustness to speech understanding errors. Various approximations allow such a model to be used for building real-world dialog systems. However, they require a large number of dialogs to train the dialog policy and hence they typically rely on the availability of a user simulator. They also require significant designer effort to hand-craft the policy representation. We investigate the use of Gaussian processes (GPs) in policy modeling to overcome these problems. We show that GP policy optimization can be implemented for a real world POMDP dialog manager, and in particular: 1) we examine different formulations of a GP policy to minimize variability in the learning process; 2) we find that the use of GP increases the learning rate by an order of magnitude thereby allowing learning by direct interaction with human users; and 3) we demonstrate that designer effort can be substantially reduced by basing the policy directly on the full belief space thereby avoiding ad hoc feature space modeling. Overall, the GP approach represents an important step forward towards fully automatic dialog policy optimization in real world systems. © 2013 IEEE.

20 Gb/s mode-group-division multiplexing employing Hermite-Gaussian launches over worst-case multimode fiber links

For the first time, mode group division multiplexing is achieved in a multimode fiber link using a 2-D Hermite-Gaussian mode launch. 20 Gb/s error-free transmission is achieved over a 250 m worst-case OM1 multimode fiber link. © OSA 2014.

Effects of the "all-fish" growth hormone transgene expression on non-specific immune functions of common carp, Cyprinus carpio L.

This study investigated non-specific immune functions of the F-2 generation of "all-fish" growth hormone transgenic carp, Cyprinus carpio L. Lysozyme activity was 145.0 (+/- 30.7) U ml(-1) in the transgenic fish serum and 105.0 (+/- 38.7) U ml(-1) in age-matched non-transgenic control fish serum, a significant difference (P < 0.01). The serum bactericidal activity in the transgenics was significantly higher than that in the controls (P < 0.05), with the percentage serum killing of 59.5% (6.83%) and 50.8% (8.67%), respectively. Values for leukocrit and phagocytic percent of macrophages in head kidney were higher in transgenics than controls (P < 0.05). However, the phagocytic indices in the transgenics and the controls were not different. In addition, the mean body weight of the transgenics was 63.4 (6.65) g, much higher than that of the controls [39.2 (+/- 3.30) g, P < 0.01]. The absolute weight of spleen of the transgenics [0.13 (+/- 0.03) g] was higher than that of the controls [0.08 (+/- 0.02) g, P < 0.01]. However, there was no difference in the relative weight of spleen between the transgenics and the controls, with the spleen mass index being 0.21% (+/- 0.02%) and 0.20% (+/- 0.03%), respectively. This study suggests that the "all-fish" growth hormone transgene expression could stimulate not only the growth but also the non-specific immune functions of carp. (c) 2006 Published by Elsevier B.V.

Bayesian inference and learning in Gaussian process state-space models with Particle MCMC

State-space models are successfully used in many areas of science, engineering and economics to model time series and dynamical systems. We present a fully Bayesian approach to inference and learning (i.e. state estimation and system identification) in nonlinear nonparametric state-space models. We place a Gaussian process prior over the state transition dynamics, resulting in a flexible model able to capture complex dynamical phenomena. To enable efficient inference, we marginalize over the transition dynamics function and, instead, infer directly the joint smoothing distribution using specially tailored Particle Markov Chain Monte Carlo samplers. Once a sample from the smoothing distribution is computed, the state transition predictive distribution can be formulated analytically. Our approach preserves the full nonparametric expressivity of the model and can make use of sparse Gaussian processes to greatly reduce computational complexity.

Effects of intraperitoneal injection of cortisol on nonspecific immune functions of Ctenopharyngodon idella

After grass carps Ctenopharyngodon idellus were injected with cortisol, with (CBC) and without (C) a cocoa butter carrier, the effects of both slowly and rapidly acting exogenous cortisol oil their non-specific immune functions were investigated. On the one hand, after injection with CBC, the cortisol concentration and lysozyme activity in fish serum were enhanced and were sustained at high levels for a long period (30 days). The killing activity in the serum declined with time, and phagocytosis of head kidney macrophages diminished significantly (P < 0.05 or P < 0.01). The leukocrit values in the high dose group (31-8 mg cortisol fish(-1)) increased over time, however, with the maximum average being 5.6% at day 30. The spleen mass index in the high dose group was 0.93 x 10(-3) after 30 days, notably lower (P < 0.05) than that in the control group. In addition, a decrease in resistance to Aeronionas hydrophilo infection in cortisol-treated fish was shown, with the final cumulative mortalities being 54.5 and 66.7% in the low and high dose groups, respectively. On the other hand, there was a decrease in both serum cortisol concentration and lysozyme activity of the experimental fish within 2 weeks after injection with C, where plasma bactericidal activities in the high dose group (31-8 mg cortisol fish(-1)) were remarkably lower (P < 0.01) than those in the control group at each sampling, but were increased slightly over time. The results of which were different from those in the CBC trial. Phagocytic activity of head kidney macrophages and spleen mass index decreased significantly (P < 0.05), while there were increases in leukocrit value and cumulative mortality due to A. hydrophila. The results of which were similar to those in the CBC trial. This study indicated that the injection of cortisol depressed the non-specific immune functions of the grass carp and increased its susceptibility to disease. (c) 2005 The Fisheries Society of the British Isles.

Classification using log Gaussian Cox processes

McCullagh and Yang (2006) suggest a family of classification algorithms based on Cox processes. We further investigate the log Gaussian variant which has a number of appealing properties. Conditioned on the covariates, the distribution over labels is given by a type of conditional Markov random field. In the supervised case, computation of the predictive probability of a single test point scales linearly with the number of training points and the multiclass generalization is straightforward. We show new links between the supervised method and classical nonparametric methods. We give a detailed analysis of the pairwise graph representable Markov random field, which we use to extend the model to semi-supervised learning problems, and propose an inference method based on graph min-cuts. We give the first experimental analysis on supervised and semi-supervised datasets and show good empirical performance.

Predictive entropy search for efficient global optimization of black-box functions

We propose a novel information-theoretic approach for Bayesian optimization called Predictive Entropy Search (PES). At each iteration, PES selects the next evaluation point that maximizes the expected information gained with respect to the global maximum. PES codifies this intractable acquisition function in terms of the expected reduction in the differential entropy of the predictive distribution. This reformulation allows PES to obtain approximations that are both more accurate and efficient than other alternatives such as Entropy Search (ES). Furthermore, PES can easily perform a fully Bayesian treatment of the model hyperparameters while ES cannot. We evaluate PES in both synthetic and real-world applications, including optimization problems in machine learning, finance, biotechnology, and robotics. We show that the increased accuracy of PES leads to significant gains in optimization performance.

Silica-based 64-channel arrayed waveguide gratings with double functions

Silica-based 64-channel arrayed waveguide gratings (AWGs) with double functions and 0.4 nm (50 GHz) channel spacing have been designed and fabricated. On the same component, Gauss and flat-top output response spectra are obtained simultaneously. The test results show that when the insertion loss ranges from 3.5 dB to 6 dB,the crosstalk is better than -34 dB, the 1 dB bandwidth is 0.12 nm, the 3 dB bandwidth is 0,218 nm, and the polarization-dependent loss (PDL) is less than 0.5 dB for Gauss response. When the insertion loss ranges,from 5.8 dB to 7.8 dB, the crosstalk is better than -30 dB, the 1 dB bandwidth is 0.24 nm, the 3 dB bandwidth is 0.33 nm, and the PDL is less than 0.2 dB for flat-top response.

A criterion for local interconvertibility between all-tripartite Gaussian states

We arrive at a necessary and sufficient criterion that can be readily used for interconvertibility between general, all-tripartite Gaussian states under local quantum operation. The derivation involves a systematic reduction that converts the original complex conditions in high-dimensional, 6n x 6n matrix space eventually into 2 x 2 matrix problems.

Conditions for the local manipulation of all bipartite Gaussian states

We provide a general, necessary, and sufficient condition for the possibility of transforming a mixed bipartite Gaussian state with arbitrarily many modes to another one under arbitrary local Gaussian channels, which do not include classical communication. Moreover, by means of this condition we present a necessary criterion that can be used to check the possibility of a state transformation between two mixed Gaussian states. At the same time, we prove that our criterion can be reduced to the Eisert-Plenio criterion when the mode number is chosen as 1 per side.