Correlates of depression in bipolar disorder

We analyse time series from 100 patients with bipolar disorder for correlates of depression symptoms. As the sampling interval is non-uniform, we quantify the extent of missing and irregular data using new measures of compliance and continuity. We find that uniformity of response is negatively correlated with the standard deviation of sleep ratings (ρ = -0.26, p = 0.01). To investigate the correlation structure of the time series themselves, we apply the Edelson-Krolik method for correlation estimation. We examine the correlation between depression symptoms for a subset of patients and find that self-reported measures of sleep and appetite/weight show a lower average correlation than other symptoms. Using surrogate time series as a reference dataset, we find no evidence that depression is correlated between patients, though we note a possible loss of information from sparse sampling. © 2013 The Author(s).

Neural network regression with input uncertainty

It is generally assumed when using Bayesian inference methods for neural networks that the input data contains no noise or corruption. For real-world (errors in variable) problems this is clearly an unsafe assumption. This paper presents a Bayesian neural network framework which allows for input noise given that some model of the noise process exists. In the limit where this noise process is small and symmetric it is shown, using the Laplace approximation, that there is an additional term to the usual Bayesian error bar which depends on the variance of the input noise process. Further, by treating the true (noiseless) input as a hidden variable and sampling this jointly with the network's weights, using Markov Chain Monte Carlo methods, it is demonstrated that it is possible to infer the unbiassed regression over the noiseless input.

Neural Network Modelling with Input Uncertainty: Theory and Application

It is generally assumed when using Bayesian inference methods for neural networks that the input data contains no noise. For real-world (errors in variable) problems this is clearly an unsafe assumption. This paper presents a Bayesian neural network framework which accounts for input noise provided that a model of the noise process exists. In the limit where the noise process is small and symmetric it is shown, using the Laplace approximation, that this method adds an extra term to the usual Bayesian error bar which depends on the variance of the input noise process. Further, by treating the true (noiseless) input as a hidden variable, and sampling this jointly with the network’s weights, using a Markov chain Monte Carlo method, it is demonstrated that it is possible to infer the regression over the noiseless input. This leads to the possibility of training an accurate model of a system using less accurate, or more uncertain, data. This is demonstrated on both the, synthetic, noisy sine wave problem and a real problem of inferring the forward model for a satellite radar backscatter system used to predict sea surface wind vectors.

Full-field electronic dispersion compensation of a 10 Gbit/s OOK signal over 4 x 124 km field-installed single-mode fibre

We experimentally demonstrate the use of full-field electronic dispersion compensation (EDC) to achieve a bit error rate of 5 x 10(-5) at 22.3 dB optical signal-to-noise ratio for single-channel 10 Gbit/s on-off keyed signal after transmission over 496 km field-installed single-mode fibre with an amplifier spacing of 124 km. This performance is achieved by designing the EDC so as to avoid electronic amplification of the noise content of the signal during full-field reconstruction. We also investigate the tolerance of the system to key signal processing parameters, and numerically demonstrate that single-channel 2160 km single mode fibre transmission without in-line optical dispersion compensation can be achieved using this technique with 80 km amplifier spacing and optimized system parameters.

In situ X-ray studies of crotyl alcohol selective oxidation over Au/Pd(1 1 1) surface alloys

The selective oxidation of crotyl alcohol to crotonaldehyde over ultrathin Au overlayers on Pd(1 1 1) and Au/Pd(1 1 1) surface alloys has been investigated by time-resolved X-ray photoelectron spectroscopy (XPS) and mass spectrometry. Pure gold is catalytically inert towards crotyl alcohol which undergoes reversible adsorption. In contrast, thermal processing of a 3.9 monolayer (ML) gold overlayer allows access to a range of AuPd surface alloy compositions, which are extremely selective towards crotonaldehyde production, and greatly reduce the extent of hydrocarbon decomposition and eventual carbon laydown compared with base Pd(1 1 1). XPS and CO titrations suggest that palladium-rich surface alloys offer the optimal balance between alcohol oxidative dehydrogenation activity while minimising competitive decomposition pathways, and that Pd monomers are not the active surface ensemble for such selox chemistry over AuPd alloys. Crown Copyright © 2008.

Unrepeatered 8 x 40Gb/s transmission over 320km SMF-28 using ultra-long raman fibre laser based amplification

Unrepeatered transmission over SMF-28 fibre is investigated using ultra-long Raman fibre laser based amplification. Experiments and simulations demonstrate 8 x 42.7Gb/s transmission up to 320km (67dB) span length using DPSK and ASK modulation with direct detection. © 2012 OSA.