Prediction with Gaussian processes: from linear regression to linear prediction and beyond

The main aim of this paper is to provide a tutorial on regression with Gaussian processes. We start from Bayesian linear regression, and show how by a change of viewpoint one can see this method as a Gaussian process predictor based on priors over functions, rather than on priors over parameters. This leads in to a more general discussion of Gaussian processes in section 4. Section 5 deals with further issues, including hierarchical modelling and the setting of the parameters that control the Gaussian process, the covariance functions for neural network models and the use of Gaussian processes in classification problems.

Bayesian inference for wind field retrieval

In many problems in spatial statistics it is necessary to infer a global problem solution by combining local models. A principled approach to this problem is to develop a global probabilistic model for the relationships between local variables and to use this as the prior in a Bayesian inference procedure. We show how a Gaussian process with hyper-parameters estimated from Numerical Weather Prediction Models yields meteorologically convincing wind fields. We use neural networks to make local estimates of wind vector probabilities. The resulting inference problem cannot be solved analytically, but Markov Chain Monte Carlo methods allow us to retrieve accurate wind fields.

Bayesian inference for wind field retrieval

In many problems in spatial statistics it is necessary to infer a global problem solution by combining local models. A principled approach to this problem is to develop a global probabilistic model for the relationships between local variables and to use this as the prior in a Bayesian inference procedure. We show how a Gaussian process with hyper-parameters estimated from Numerical Weather Prediction Models yields meteorologically convincing wind fields. We use neural networks to make local estimates of wind vector probabilities. The resulting inference problem cannot be solved analytically, but Markov Chain Monte Carlo methods allow us to retrieve accurate wind fields.

A pragmatic Bayesian approach to wind field retrieval

The ERS-1 Satellite was launched in July 1991 by the European Space Agency into a polar orbit at about 800 km, carrying a C-band scatterometer. A scatterometer measures the amount of backscatter microwave radiation reflected by small ripples on the ocean surface induced by sea-surface winds, and so provides instantaneous snap-shots of wind flow over large areas of the ocean surface, known as wind fields. Inherent in the physics of the observation process is an ambiguity in wind direction; the scatterometer cannot distinguish if the wind is blowing toward or away from the sensor device. This ambiguity implies that there is a one-to-many mapping between scatterometer data and wind direction. Current operational methods for wind field retrieval are based on the retrieval of wind vectors from satellite scatterometer data, followed by a disambiguation and filtering process that is reliant on numerical weather prediction models. The wind vectors are retrieved by the local inversion of a forward model, mapping scatterometer observations to wind vectors, and minimising a cost function in scatterometer measurement space. This thesis applies a pragmatic Bayesian solution to the problem. The likelihood is a combination of conditional probability distributions for the local wind vectors given the scatterometer data. The prior distribution is a vector Gaussian process that provides the geophysical consistency for the wind field. The wind vectors are retrieved directly from the scatterometer data by using mixture density networks, a principled method to model multi-modal conditional probability density functions. The complexity of the mapping and the structure of the conditional probability density function are investigated. A hybrid mixture density network, that incorporates the knowledge that the conditional probability distribution of the observation process is predominantly bi-modal, is developed. The optimal model, which generalises across a swathe of scatterometer readings, is better on key performance measures than the current operational model. Wind field retrieval is approached from three perspectives. The first is a non-autonomous method that confirms the validity of the model by retrieving the correct wind field 99% of the time from a test set of 575 wind fields. The second technique takes the maximum a posteriori probability wind field retrieved from the posterior distribution as the prediction. For the third technique, Markov Chain Monte Carlo (MCMC) techniques were employed to estimate the mass associated with significant modes of the posterior distribution, and make predictions based on the mode with the greatest mass associated with it. General methods for sampling from multi-modal distributions were benchmarked against a specific MCMC transition kernel designed for this problem. It was shown that the general methods were unsuitable for this application due to computational expense. On a test set of 100 wind fields the MAP estimate correctly retrieved 72 wind fields, whilst the sampling method correctly retrieved 73 wind fields.

Bayesian data assimilation

This thesis addresses data assimilation, which typically refers to the estimation of the state of a physical system given a model and observations, and its application to short-term precipitation forecasting. A general introduction to data assimilation is given, both from a deterministic and' stochastic point of view. Data assimilation algorithms are reviewed, in the static case (when no dynamics are involved), then in the dynamic case. A double experiment on two non-linear models, the Lorenz 63 and the Lorenz 96 models, is run and the comparative performance of the methods is discussed in terms of quality of the assimilation, robustness "in the non-linear regime and computational time. Following the general review and analysis, data assimilation is discussed in the particular context of very short-term rainfall forecasting (nowcasting) using radar images. An extended Bayesian precipitation nowcasting model is introduced. The model is stochastic in nature and relies on the spatial decomposition of the rainfall field into rain "cells". Radar observations are assimilated using a Variational Bayesian method in which the true posterior distribution of the parameters is approximated by a more tractable distribution. The motion of the cells is captured by a 20 Gaussian process. The model is tested on two precipitation events, the first dominated by convective showers, the second by precipitation fronts. Several deterministic and probabilistic validation methods are applied and the model is shown to retain reasonable prediction skill at up to 3 hours lead time. Extensions to the model are discussed.

Toward bacterial protein sub-cellular location prediction:single-class discrimminant models for all gram- and gram+ compartments

Based on Bayesian Networks, methods were created that address protein sequence-based bacterial subcellular location prediction. Distinct predictive algorithms for the eight bacterial subcellular locations were created. Several variant methods were explored. These variations included differences in the number of residues considered within the query sequence - which ranged from the N-terminal 10 residues to the whole sequence - and residue representation - which took the form of amino acid composition, percentage amino acid composition, or normalised amino acid composition. The accuracies of the best performing networks were then compared to PSORTB. All individual location methods outperform PSORTB except for the Gram+ cytoplasmic protein predictor, for which accuracies were essentially equal, and for outer membrane protein prediction, where PSORTB outperforms the binary predictor. The method described here is an important new approach to method development for subcellular location prediction. It is also a new, potentially valuable tool for candidate subunit vaccine selection.

TATPred:a Bayesian method for the identification of twin arginine translocation pathway signal sequences

The twin arginine translocation (TAT) system ferries folded proteins across the bacterial membrane. Proteins are directed into this system by the TAT signal peptide present at the amino terminus of the precursor protein, which contains the twin arginine residues that give the system its name. There are currently only two computational methods for the prediction of TAT translocated proteins from sequence. Both methods have limitations that make the creation of a new algorithm for TAT-translocated protein prediction desirable. We have developed TATPred, a new sequence-model method, based on a Nave-Bayesian network, for the prediction of TAT signal peptides. In this approach, a comprehensive range of models was tested to identify the most reliable and robust predictor. The best model comprised 12 residues: three residues prior to the twin arginines and the seven residues that follow them. We found a prediction sensitivity of 0.979 and a specificity of 0.942.

LIPPRED:a web server for accurate prediction of lipoprotein signal sequences and cleavage sites

Bacterial lipoproteins have many important functions and represent a class of possible vaccine candidates. The prediction of lipoproteins from sequence is thus an important task for computational vaccinology. Naïve-Bayesian networks were trained to identify SpaseII cleavage sites and their preceding signal sequences using a set of 199 distinct lipoprotein sequences. A comprehensive range of sequence models was used to identify the best model for lipoprotein signal sequences. The best performing sequence model was found to be 10-residues in length, including the conserved cysteine lipid attachment site and the nine residues prior to it. The sensitivity of prediction for LipPred was 0.979, while the specificity was 0.742. Here, we describe LipPred, a web server for lipoprotein prediction; available at the URL: http://www.jenner.ac.uk/LipPred/. LipPred is the most accurate method available for the detection of SpaseIIcleaved lipoprotein signal sequences and the prediction of their cleavage sites.

A Note on Bayesian Estimation for the Negative-Binomial Model

2000 Mathematics Subject Classification: 62F15.

Seeking Relationships in Big Data: A Bayesian Perspective

The real purpose of collecting big data is to identify causality in the hope that this will facilitate credible predictivity . But the search for causality can trap one into infinite regress, and thus one takes refuge in seeking associations between variables in data sets. Regrettably, the mere knowledge of associations does not enable predictivity. Associations need to be embedded within the framework of probability calculus to make coherent predictions. This is so because associations are a feature of probability models, and hence they do not exist outside the framework of a model. Measures of association, like correlation, regression, and mutual information merely refute a preconceived model. Estimated measures of associations do not lead to a probability model; a model is the product of pure thought. This paper discusses these and other fundamentals that are germane to seeking associations in particular, and machine learning in general. ACM Computing Classification System (1998): H.1.2, H.2.4., G.3.

Handling high level of censoring for endovascular aortic repair risk prediction

Feature selection is important in medical field for many reasons. However, selecting important variables is a difficult task with the presence of censoring that is a unique feature in survival data analysis. This paper proposed an approach to deal with the censoring problem in endovascular aortic repair survival data through Bayesian networks. It was merged and embedded with a hybrid feature selection process that combines cox's univariate analysis with machine learning approaches such as ensemble artificial neural networks to select the most relevant predictive variables. The proposed algorithm was compared with common survival variable selection approaches such as; least absolute shrinkage and selection operator LASSO, and Akaike information criterion AIC methods. The results showed that it was capable of dealing with high censoring in the datasets. Moreover, ensemble classifiers increased the area under the roc curves of the two datasets collected from two centers located in United Kingdom separately. Furthermore, ensembles constructed with center 1 enhanced the concordance index of center 2 prediction compared to the model built with a single network. Although the size of the final reduced model using the neural networks and its ensembles is greater than other methods, the model outperformed the others in both concordance index and sensitivity for center 2 prediction. This indicates the reduced model is more powerful for cross center prediction.

A Bayesian neural network algorithm identifies patients in whom post-EVAR surveillance may be unnecessary

Lifelong surveillance is not cost-effective after endovascular aneurysm repair (EVAR), but is required to detect aortic complications which are fatal if untreated (type 1/3 endoleak, sac expansion, device migration). Aneurysm morphology determines the probability of aortic complications and therefore the need for surveillance, but existing analyses have proven incapable of identifying patients at sufficiently low risk to justify abandoning surveillance. This study aimed to improve the prediction of aortic complications, through the application of machine-learning techniques. Patients undergoing EVAR at 2 centres were studied from 2004–2010. Aneurysm morphology had previously been studied to derive the SGVI Score for predicting aortic complications. Bayesian Neural Networks were designed using the same data, to dichotomise patients into groups at low- or high-risk of aortic complications. Network training was performed only on patients treated at centre 1. External validation was performed by assessing network performance independently of network training, on patients treated at centre 2. Discrimination was assessed by Kaplan-Meier analysis to compare aortic complications in predicted low-risk versus predicted high-risk patients. 761 patients aged 75 +/− 7 years underwent EVAR in 2 centres. Mean follow-up was 36+/− 20 months. Neural networks were created incorporating neck angu- lation/length/diameter/volume; AAA diameter/area/volume/length/tortuosity; and common iliac tortuosity/diameter. A 19-feature network predicted aor- tic complications with excellent discrimination and external validation (5-year freedom from aortic complications in predicted low-risk vs predicted high-risk patients: 97.9% vs. 63%; p < 0.0001). A Bayesian Neural-Network algorithm can identify patients in whom it may be safe to abandon surveillance after EVAR. This proposal requires prospective study.

Proposing the deep dynamic Bayesian network as a future computer based medical system

The development of new learning models has been of great importance throughout recent years, with a focus on creating advances in the area of deep learning. Deep learning was first noted in 2006, and has since become a major area of research in a number of disciplines. This paper will delve into the area of deep learning to present its current limitations and provide a new idea for a fully integrated deep and dynamic probabilistic system. The new model will be applicable to a vast number of areas initially focusing on applications into medical image analysis with an overall goal of utilising this approach for prediction purposes in computer based medical systems.

Factor analysis applied to genome prediction for high-dimensional phenotypes in pigs.

The aim of the present study was to propose and evaluate the use of factor analysis (FA) in obtaining latent variables (factors) that represent a set of pig traits simultaneously, for use in genome-wide selection (GWS) studies. We used crosses between outbred F2 populations of Brazilian Piau X commercial pigs. Data were obtained on 345 F2 pigs, genotyped for 237 SNPs, with 41 traits. FA allowed us to obtain four biologically interpretable factors: ?weight?, ?fat?, ?loin?, and ?performance?. These factors were used as dependent variables in multiple regression models of genomic selection (Bayes A, Bayes B, RR-BLUP, and Bayesian LASSO). The use of FA is presented as an interesting alternative to select individuals for multiple variables simultaneously in GWS studies; accuracy measurements of the factors were similar to those obtained when the original traits were considered individually. The similarities between the top 10% of individuals selected by the factor, and those selected by the individual traits, were also satisfactory. Moreover, the estimated markers effects for the traits were similar to those found for the relevant factor.

Petroleomics By Electrospray Ionization Ft-icr Mass Spectrometry Coupled To Partial Least Squares With Variable Selection Methods: Prediction Of The Total Acid Number Of Crude Oils.

Negative-ion mode electrospray ionization, ESI(-), with Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) was coupled to a Partial Least Squares (PLS) regression and variable selection methods to estimate the total acid number (TAN) of Brazilian crude oil samples. Generally, ESI(-)-FT-ICR mass spectra present a power of resolution of ca. 500,000 and a mass accuracy less than 1 ppm, producing a data matrix containing over 5700 variables per sample. These variables correspond to heteroatom-containing species detected as deprotonated molecules, [M - H](-) ions, which are identified primarily as naphthenic acids, phenols and carbazole analog species. The TAN values for all samples ranged from 0.06 to 3.61 mg of KOH g(-1). To facilitate the spectral interpretation, three methods of variable selection were studied: variable importance in the projection (VIP), interval partial least squares (iPLS) and elimination of uninformative variables (UVE). The UVE method seems to be more appropriate for selecting important variables, reducing the dimension of the variables to 183 and producing a root mean square error of prediction of 0.32 mg of KOH g(-1). By reducing the size of the data, it was possible to relate the selected variables with their corresponding molecular formulas, thus identifying the main chemical species responsible for the TAN values.