Reduced dimension models for weather forecasting

DUE TO COPYRIGHT RESTRICTIONS ONLY AVAILABLE FOR CONSULTATION AT ASTON UNIVERSITY LIBRARY AND INFORMATION SERVICES WITH PRIOR ARRANGEMENT

Projected sequential Gaussian processes:a C++ tool for interpolation of large datasets with heterogeneous noise

Heterogeneous datasets arise naturally in most applications due to the use of a variety of sensors and measuring platforms. Such datasets can be heterogeneous in terms of the error characteristics and sensor models. Treating such data is most naturally accomplished using a Bayesian or model-based geostatistical approach; however, such methods generally scale rather badly with the size of dataset, and require computationally expensive Monte Carlo based inference. Recently within the machine learning and spatial statistics communities many papers have explored the potential of reduced rank representations of the covariance matrix, often referred to as projected or fixed rank approaches. In such methods the covariance function of the posterior process is represented by a reduced rank approximation which is chosen such that there is minimal information loss. In this paper a sequential Bayesian framework for inference in such projected processes is presented. The observations are considered one at a time which avoids the need for high dimensional integrals typically required in a Bayesian approach. A C++ library, gptk, which is part of the INTAMAP web service, is introduced which implements projected, sequential estimation and adds several novel features. In particular the library includes the ability to use a generic observation operator, or sensor model, to permit data fusion. It is also possible to cope with a range of observation error characteristics, including non-Gaussian observation errors. Inference for the covariance parameters is explored, including the impact of the projected process approximation on likelihood profiles. We illustrate the projected sequential method in application to synthetic and real datasets. Limitations and extensions are discussed. © 2010 Elsevier Ltd.

Mathematical modelling and control of agitated extraction columns

Liquid-liquid extraction has long been known as a unit operation that plays an important role in industry. This process is well known for its complexity and sensitivity to operation conditions. This thesis presents an attempt to explore the dynamics and control of this process using a systematic approach and state of the art control system design techniques. The process was studied first experimentally under carefully selected. operation conditions, which resembles the ranges employed practically under stable and efficient conditions. Data were collected at steady state conditions using adequate sampling techniques for the dispersed and continuous phases as well as during the transients of the column with the aid of a computer-based online data logging system and online concentration analysis. A stagewise single stage backflow model was improved to mimic the dynamic operation of the column. The developed model accounts for the variation in hydrodynamics, mass transfer, and physical properties throughout the length of the column. End effects were treated by addition of stages at the column entrances. Two parameters were incorporated in the model namely; mass transfer weight factor to correct for the assumption of no mass transfer in the. settling zones at each stage and the backmixing coefficients to handle the axial dispersion phenomena encountered in the course of column operation. The parameters were estimated by minimizing the differences between the experimental and the model predicted concentration profiles at steady state conditions using non-linear optimisation technique. The estimated values were then correlated as functions of operating parameters and were incorporated in·the model equations. The model equations comprise a stiff differential~algebraic system. This system was solved using the GEAR ODE solver. The calculated concentration profiles were compared to those experimentally measured. A very good agreement of the two profiles was achieved within a percent relative error of ±2.S%. The developed rigorous dynamic model of the extraction column was used to derive linear time-invariant reduced-order models that relate the input variables (agitator speed, solvent feed flowrate and concentration, feed concentration and flowrate) to the output variables (raffinate concentration and extract concentration) using the asymptotic method of system identification. The reduced-order models were shown to be accurate in capturing the dynamic behaviour of the process with a maximum modelling prediction error of I %. The simplicity and accuracy of the derived reduced-order models allow for control system design and analysis of such complicated processes. The extraction column is a typical multivariable process with agitator speed and solvent feed flowrate considered as manipulative variables; raffinate concentration and extract concentration as controlled variables and the feeds concentration and feed flowrate as disturbance variables. The control system design of the extraction process was tackled as multi-loop decentralised SISO (Single Input Single Output) as well as centralised MIMO (Multi-Input Multi-Output) system using both conventional and model-based control techniques such as IMC (Internal Model Control) and MPC (Model Predictive Control). Control performance of each control scheme was. studied in terms of stability, speed of response, sensitivity to modelling errors (robustness), setpoint tracking capabilities and load rejection. For decentralised control, multiple loops were assigned to pair.each manipulated variable with each controlled variable according to the interaction analysis and other pairing criteria such as relative gain array (RGA), singular value analysis (SVD). Loops namely Rotor speed-Raffinate concentration and Solvent flowrate Extract concentration showed weak interaction. Multivariable MPC has shown more effective performance compared to other conventional techniques since it accounts for loops interaction, time delays, and input-output variables constraints.

Projected sequential Gaussian processes: flexible interpolation for large data sets

Recently within the machine learning and spatial statistics communities many papers have explored the potential of reduced rank representations of the covariance matrix, often referred to as projected or fixed rank approaches. In such methods the covariance function of the posterior process is represented by a reduced rank approximation which is chosen such that there is minimal information loss. In this paper a sequential framework for inference in such projected processes is presented, where the observations are considered one at a time. We introduce a C++ library for carrying out such projected, sequential estimation which adds several novel features. In particular we have incorporated the ability to use a generic observation operator, or sensor model, to permit data fusion. We can also cope with a range of observation error characteristics, including non-Gaussian observation errors. Inference for the variogram parameters is based on maximum likelihood estimation. We illustrate the projected sequential method in application to synthetic and real data sets. We discuss the software implementation and suggest possible future extensions.

Instantiating deformable models with a neural net

Deformable models are an attractive approach to recognizing objects which have considerable within-class variability such as handwritten characters. However, there are severe search problems associated with fitting the models to data which could be reduced if a better starting point for the search were available. We show that by training a neural network to predict how a deformable model should be instantiated from an input image, such improved starting points can be obtained. This method has been implemented for a system that recognizes handwritten digits using deformable models, and the results show that the search time can be significantly reduced without compromising recognition performance. © 1997 Academic Press.

Forecasting exchange rates with linear and nonlinear models

In this paper, the exchange rate forecasting performance of neural network models are evaluated against the random walk, autoregressive moving average and generalised autoregressive conditional heteroskedasticity models. There are no guidelines available that can be used to choose the parameters of neural network models and therefore, the parameters are chosen according to what the researcher considers to be the best. Such an approach, however,implies that the risk of making bad decisions is extremely high, which could explain why in many studies, neural network models do not consistently perform better than their time series counterparts. In this paper, through extensive experimentation, the level of subjectivity in building neural network models is considerably reduced and therefore giving them a better chance of Forecasting exchange rates with linear and nonlinear models 415 performing well. The results show that in general, neural network models perform better than the traditionally used time series models in forecasting exchange rates.

Added luminance ramp alters perceived edge blur and contrast:A critical test for derivative-based models of edge coding

In many models of edge analysis in biological vision, the initial stage is a linear 2nd derivative operation. Such models predict that adding a linear luminance ramp to an edge will have no effect on the edge's appearance, since the ramp has no effect on the 2nd derivative. Our experiments did not support this prediction: adding a negative-going ramp to a positive-going edge (or vice-versa) greatly reduced the perceived blur and contrast of the edge. The effects on a fairly sharp edge were accurately predicted by a nonlinear multi-scale model of edge processing [Georgeson, M. A., May, K. A., Freeman, T. C. A., & Hesse, G. S. (in press). From filters to features: Scale-space analysis of edge and blur coding in human vision. Journal of Vision], in which a half-wave rectifier comes after the 1st derivative filter. But we also found that the ramp affected perceived blur more profoundly when the edge blur was large, and this greater effect was not predicted by the existing model. The model's fit to these data was much improved when the simple half-wave rectifier was replaced by a threshold-like transducer [May, K. A. & Georgeson, M. A. (2007). Blurred edges look faint, and faint edges look sharp: The effect of a gradient threshold in a multi-scale edge coding model. Vision Research, 47, 1705-1720.]. This modified model correctly predicted that the interaction between ramp gradient and edge scale would be much larger for blur perception than for contrast perception. In our model, the ramp narrows an internal representation of the gradient profile, leading to a reduction in perceived blur. This in turn reduces perceived contrast because estimated blur plays a role in the model's estimation of contrast. Interestingly, the model predicts that analogous effects should occur when the width of the window containing the edge is made narrower. This has already been confirmed for blur perception; here, we further support the model by showing a similar effect for contrast perception. © 2007 Elsevier Ltd. All rights reserved.

Updating finite element models using measured vibration data

The modelling of mechanical structures using finite element analysis has become an indispensable stage in the design of new components and products. Once the theoretical design has been optimised a prototype may be constructed and tested. What can the engineer do if the measured and theoretically predicted vibration characteristics of the structure are significantly different? This thesis considers the problems of changing the parameters of the finite element model to improve the correlation between a physical structure and its mathematical model. Two new methods are introduced to perform the systematic parameter updating. The first uses the measured modal model to derive the parameter values with the minimum variance. The user must provide estimates for the variance of the theoretical parameter values and the measured data. Previous authors using similar methods have assumed that the estimated parameters and measured modal properties are statistically independent. This will generally be the case during the first iteration but will not be the case subsequently. The second method updates the parameters directly from the frequency response functions. The order of the finite element model of the structure is reduced as a function of the unknown parameters. A method related to a weighted equation error algorithm is used to update the parameters. After each iteration the weighting changes so that on convergence the output error is minimised. The suggested methods are extensively tested using simulated data. An H frame is then used to demonstrate the algorithms on a physical structure.

Forecasting exchange rates with linear and nonlinear models

In this paper the exchange rate forecasting performance of neural network models are evaluated against random walk and a range of time series models. There are no guidelines available that can be used to choose the parameters of neural network models and therefore the parameters are chosen according to what the researcher considers to be the best. Such an approach, however, implies that the risk of making bad decisions is extremely high which could explain why in many studies neural network models do not consistently perform better than their time series counterparts. In this paper through extensive experimentation the level of subjectivity in building neural network models is considerably reduced and therefore giving them a better chance of performing well. Our results show that in general neural network models perform better than traditionally used time series models in forecasting exchange rates.

Effects of some 5-hydroxytryptamine and related ligands in anxiety models

Drugs acting at 5-HT receptors were evaluated on three animal models of anxiety. On the elevated X-maze test the majority of 5-HT1 agonists were found to be anxiogenic. However, ipsapirone was anxiolytic and buspirone and gepirone were inactive. The 5-HT2 agonist DOI and the 5-HT2 antagonist ritanserin were anxiolytic while ICI 169,369, a 5-HT2 antagonist was inactive. All 5-HT3 antagonists tested were inactive in this test, while the indirect serotomimetics zimeldine and fenfluramine were anxiogenic. Neither beta-adrenoceptor agonists nor antagonists had reproducible effects on anxiety in this model. Combined beta-1/beta-2 adrenoceptor antagonists reversed the anxiogenic effects of 8-OH-DPAT while selective beta-1 or beta-2 antagonists did not. On the social interaction model the 5-HT1 agonists 8-OH-DPAT, RU 24969 and 5-MeODMT were anxiogenic and ipsapirone was anxiolytic. The 5-HT2 agonist DOI and the beta-adrenoceptor- and 5-HT- antagonist pindolol were anxiolytic, while the 5-HT2 and 5-HT3 antagonists were inactive. In the marble burying test, the 5-HT upake inhibitors zimeldine, fluvoxamine, indalpine and citalopram, the 5-HT1B/5-HT1C agonists mCPP and TFMPP and the 5-HT2/5-HT1C agonist DOI reduced marble burying without affecting locomotor activity. 5-HT1A agonists and the 5-HT2 and 5-HT3 antagonists were without effect. Lesions of the dorsal raphe nucleus reversed the anxiogenic effects of 8-OH-DPAT in the X-maze model. The implication of these results for the understanding of the pharmacology of 5-HT in anxiety is discussed.

Multi-level visualisation using Gaussian process latent variable models

Projection of a high-dimensional dataset onto a two-dimensional space is a useful tool to visualise structures and relationships in the dataset. However, a single two-dimensional visualisation may not display all the intrinsic structure. Therefore, hierarchical/multi-level visualisation methods have been used to extract more detailed understanding of the data. Here we propose a multi-level Gaussian process latent variable model (MLGPLVM). MLGPLVM works by segmenting data (with e.g. K-means, Gaussian mixture model or interactive clustering) in the visualisation space and then fitting a visualisation model to each subset. To measure the quality of multi-level visualisation (with respect to parent and child models), metrics such as trustworthiness, continuity, mean relative rank errors, visualisation distance distortion and the negative log-likelihood per point are used. We evaluate the MLGPLVM approach on the ‘Oil Flow’ dataset and a dataset of protein electrostatic potentials for the ‘Major Histocompatibility Complex (MHC) class I’ of humans. In both cases, visual observation and the quantitative quality measures have shown better visualisation at lower levels.

Comparison of the anticatabolic effects of leucine and Ca-β-hydroxy-β-methylbutyrate in experimental models of cancer cachexia

Objective: Loss of skeletal muscle is the most debilitating feature of cancer cachexia, and there are few treatments available. The aim of this study was to compare the anticatabolic efficacy of L-leucine and the leucine metabolite β-hydroxy-β-methylbutyrate (Ca-HMB) on muscle protein metabolism, both invitro and invivo. Methods: Studies were conducted in mice bearing the cachexia-inducing murine adenocarcinoma 16 tumor, and in murine C2 C12 myotubes exposed to proteolysis-inducing factor, lipopolysaccharide, and angiotensin II. Results: Both leucine and HMB were found to attenuate the increase in protein degradation and the decrease in protein synthesis in murine myotubes induced by proteolysis-inducing factor, lipopolysaccharide, and angiotensin II. However, HMB was more potent than leucine, because HMB at 50 μM produced essentially the same effect as leucine at 1 mM. Both leucine and HMB reduced the activity of the ubiquitin-proteasome pathway as measured by the functional (chymotrypsin-like) enzyme activity of the proteasome in muscle lysates, as well as Western blot quantitation of protein levels of the structural/enzymatic proteasome subunits (20 S and 19 S) and the ubiquitin ligases (MuRF1 and MAFbx). Invivo studies in mice bearing the murine adenocarcinoma 16 tumor showed a low dose of Ca-HMB (0.25 g/kg) tobe 60% more effective than leucine (1 g/kg) in attenuating loss of body weight over a 4-d period. Conclusion: These results favor the clinical feasibility of using Ca-HMB over high doses of leucine for the treatment of cancer cachexia. © 2014 Elsevier Inc.

Detection and quantification of protein oxidation in sarcopenic models:a mass spectrometry study

Oxidised biomolecules in aged tissue could potentially be used as biomarkers for age-related diseases; however, it is still unclear whether they causatively contribute to ageing or are consequences of the ageing process. To assess the potential of using protein oxidation as markers of ageing, mass spectrometry (MS) was employed for the identification and quantification of oxidative modifications in obese (ob/ob) mice. Lean muscle mass and strength is reduced in obesity, representing a sarcopenic model in which the levels of oxidation can be evaluated for different muscular systems including calcium homeostasis, metabolism and contractility. Several oxidised residues were identified by tandem MS (MS/MS) in both muscle homogenate and isolated sarcoplasmic reticulum (SR), an organelle that regulates intracellular calcium levels in muscle. These modifications include oxidation of methionine, cysteine, tyrosine, and tryptophan in several proteins such as sarcoplasmic reticulum calcium ATPase (SERCA), glycogen phosphorylase, and myosin. Once modifications had been identified, multiple reaction monitoring MS (MRM) was used to quantify the percentage modification of oxidised residues within the samples. Preliminary data suggests proteins in ob/ob mice are more oxidised than the controls. For example SERCA, which constitutes 60-70% of the SR, had approximately a 2-fold increase in cysteine trioxidation of Cys561 in the obese model when compared to the control. Other obese muscle proteins have also shown a similar increase in oxidation for various residues. Further analysis with complex protein mixtures will determine the potential diagnostic use of MRM experiments for analysing protein oxidation in small biological samples such as muscle needle biopsies.

A low mortality, high morbidity Reduced Intensity Status Epilepticus (RISE) model of epilepsy and epileptogenesis in the rat

Animal models of acquired epilepsies aim to provide researchers with tools for use in understanding the processes underlying the acquisition, development and establishment of the disorder. Typically, following a systemic or local insult, vulnerable brain regions undergo a process leading to the development, over time, of spontaneous recurrent seizures. Many such models make use of a period of intense seizure activity or status epilepticus, and this may be associated with high mortality and/or global damage to large areas of the brain. These undesirable elements have driven improvements in the design of chronic epilepsy models, for example the lithium-pilocarpine epileptogenesis model. Here, we present an optimised model of chronic epilepsy that reduces mortality to 1% whilst retaining features of high epileptogenicity and development of spontaneous seizures. Using local field potential recordings from hippocampus in vitro as a probe, we show that the model does not result in significant loss of neuronal network function in area CA3 and, instead, subtle alterations in network dynamics appear during a process of epileptogenesis, which eventually leads to a chronic seizure state. The model’s features of very low mortality and high morbidity in the absence of global neuronal damage offer the chance to explore the processes underlying epileptogenesis in detail, in a population of animals not defined by their resistance to seizures, whilst acknowledging and being driven by the 3Rs (Replacement, Refinement and Reduction of animal use in scientific procedures) principles.

Some extended psychological benefits of challenging social stereotypes:decreased dehumanization and a reduced reliance on heuristic thinking

One way to promote equality is to encourage people to generate counterstereotypic role models. In two experiments, we demonstrate that such interventions have much broader benefits than previously thoughtreducing a reliance on heuristic thinking and decreasing tendencies to dehumanize outgroups. In Experiment 1, participants who thought about a gender counterstereotype (e.g., a female mechanic) demonstrated a generalized decrease in dehumanization towards a range of unrelated target groups (including asylum seekers and the homeless). In Experiment 2 we replicated these findings using alternative targets and measures of dehumanization. Furthermore, we found the effect was mediated by a reduced reliance on heuristic thinking. The findings suggest educational initiatives that aim to challenge social stereotypes may not only have societal benefits (generalized tolerance), but also tangible benefits for individuals (enhanced cognitive flexibility).