Urinary proteomic biomarkers in coronary artery disease

Urinary proteomics is emerging as a powerful non-invasive tool for diagnosis and monitoring of variety of human diseases. We tested whether signatures of urinary polypeptides can contribute to the existing biomarkers for coronary artery disease (CAD). We examined a total of 359 urine samples from 88 patients with severe CAD and 282 controls. Spot urine was analyzed using capillary electrophoresis on-line coupled to ESI-TOF-MS enabling characterization of more than 1000 polypeptides per sample. In a first step a "training set" for biomarker definition was created. Multiple biomarker patterns clearly distinguished healthy controls from CAD patients, and we extracted 15 peptides that define a characteristic CAD signature panel. In a second step, the ability of the CAD-specific panel to predict the presence of CAD was evaluated in a blinded study using a "test set." The signature panel showed sensitivity of 98% (95% confidence interval, 88.7-99.6) and 83% specificity (95% confidence interval, 51.6-97.4). Furthermore the peptide pattern significantly changed toward the healthy signature correlating with the level of physical activity after therapeutic intervention. Our results show that urinary proteomics can identify CAD patients with high confidence and might also play a role in monitoring the effects of therapeutic interventions. The workflow is amenable to clinical routine testing suggesting that non-invasive proteomics analysis can become a valuable addition to other biomarkers used in cardiovascular risk assessment.

Improved neural network scatterometer forward models

Current methods for retrieving near-surface winds from scatterometer observations over the ocean surface require a forward sensor model which maps the wind vector to the measured backscatter. This paper develops a hybrid neural network forward model, which retains the physical understanding embodied in CMOD4, but incorporates greater flexibility, allowing a better fit to the observations. By introducing a separate model for the midbeam and using a common model for the fore and aft beams, we show a significant improvement in local wind vector retrieval. The hybrid model also fits the scatterometer observations more closely. The model is trained in a Bayesian framework, accounting for the noise on the wind vector inputs. We show that adding more high wind speed observations in the training set improves wind vector retrieval at high wind speeds without compromising performance at medium or low wind speeds. Copyright 2001 by the American Geophysical Union.

Bayesian error bars for regression

Regression problems are concerned with predicting the values of one or more continuous quantities, given the values of a number of input variables. For virtually every application of regression, however, it is also important to have an indication of the uncertainty in the predictions. Such uncertainties are expressed in terms of the error bars, which specify the standard deviation of the distribution of predictions about the mean. Accurate estimate of error bars is of practical importance especially when safety and reliability is an issue. The Bayesian view of regression leads naturally to two contributions to the error bars. The first arises from the intrinsic noise on the target data, while the second comes from the uncertainty in the values of the model parameters which manifests itself in the finite width of the posterior distribution over the space of these parameters. The Hessian matrix which involves the second derivatives of the error function with respect to the weights is needed for implementing the Bayesian formalism in general and estimating the error bars in particular. A study of different methods for evaluating this matrix is given with special emphasis on the outer product approximation method. The contribution of the uncertainty in model parameters to the error bars is a finite data size effect, which becomes negligible as the number of data points in the training set increases. A study of this contribution is given in relation to the distribution of data in input space. It is shown that the addition of data points to the training set can only reduce the local magnitude of the error bars or leave it unchanged. Using the asymptotic limit of an infinite data set, it is shown that the error bars have an approximate relation to the density of data in input space.

Semantic sentiment analysis of Twitter

Sentiment analysis over Twitter offer organisations a fast and effective way to monitor the publics' feelings towards their brand, business, directors, etc. A wide range of features and methods for training sentiment classifiers for Twitter datasets have been researched in recent years with varying results. In this paper, we introduce a novel approach of adding semantics as additional features into the training set for sentiment analysis. For each extracted entity (e.g. iPhone) from tweets, we add its semantic concept (e.g. Apple product) as an additional feature, and measure the correlation of the representative concept with negative/positive sentiment. We apply this approach to predict sentiment for three different Twitter datasets. Our results show an average increase of F harmonic accuracy score for identifying both negative and positive sentiment of around 6.5% and 4.8% over the baselines of unigrams and part-of-speech features respectively. We also compare against an approach based on sentiment-bearing topic analysis, and find that semantic features produce better Recall and F score when classifying negative sentiment, and better Precision with lower Recall and F score in positive sentiment classification.

Identifying candidate subunit vaccines using an alignment-independent method based on principal amino acid properties

Subunit vaccine discovery is an accepted clinical priority. The empirical approach is time- and labor-consuming and can often end in failure. Rational information-driven approaches can overcome these limitations in a fast and efficient manner. However, informatics solutions require reliable algorithms for antigen identification. All known algorithms use sequence similarity to identify antigens. However, antigenicity may be encoded subtly in a sequence and may not be directly identifiable by sequence alignment. We propose a new alignment-independent method for antigen recognition based on the principal chemical properties of protein amino acid sequences. The method is tested by cross-validation on a training set of bacterial antigens and external validation on a test set of known antigens. The prediction accuracy is 83% for the cross-validation and 80% for the external test set. Our approach is accurate and robust, and provides a potent tool for the in silico discovery of medically relevant subunit vaccines.

Class I T-cell epitope prediction:improvements using a combination of proteasome cleavage, TAP affinity, and MHC binding

Cleavage by the proteasome is responsible for generating the C terminus of T-cell epitopes. Modeling the process of proteasome cleavage as part of a multi-step algorithm for T-cell epitope prediction will reduce the number of non-binders and increase the overall accuracy of the predictive algorithm. Quantitative matrix-based models for prediction of the proteasome cleavage sites in a protein were developed using a training set of 489 naturally processed T-cell epitopes (nonamer peptides) associated with HLA-A and HLA-B molecules. The models were validated using an external test set of 227 T-cell epitopes. The performance of the models was good, identifying 76% of the C-termini correctly. The best model of proteasome cleavage was incorporated as the first step in a three-step algorithm for T-cell epitope prediction, where subsequent steps predicted TAP affinity and MHC binding using previously derived models.

Subject identification via ECG fiducial-based systems:influence of the type of QT interval correction

Electrocardiography (ECG) has been recently proposed as biometric trait for identification purposes. Intra-individual variations of ECG might affect identification performance. These variations are mainly due to Heart Rate Variability (HRV). In particular, HRV causes changes in the QT intervals along the ECG waveforms. This work is aimed at analysing the influence of seven QT interval correction methods (based on population models) on the performance of ECG-fiducial-based identification systems. In addition, we have also considered the influence of training set size, classifier, classifier ensemble as well as the number of consecutive heartbeats in a majority voting scheme. The ECG signals used in this study were collected from thirty-nine subjects within the Physionet open access database. Public domain software was used for fiducial points detection. Results suggested that QT correction is indeed required to improve the performance. However, there is no clear choice among the seven explored approaches for QT correction (identification rate between 0.97 and 0.99). MultiLayer Perceptron and Support Vector Machine seemed to have better generalization capabilities, in terms of classification performance, with respect to Decision Tree-based classifiers. No such strong influence of the training-set size and the number of consecutive heartbeats has been observed on the majority voting scheme.

Neural Networks: A Diagnostic Tool for Gastric Electrical Uncoupling?

Neural Networks have been successfully employed in different biomedical settings. They have been useful for feature extractions from images and biomedical data in a variety of diagnostic applications. In this paper, they are applied as a diagnostic tool for classifying different levels of gastric electrical uncoupling in controlled acute experiments on dogs. Data was collected from 16 dogs using six bipolar electrodes inserted into the serosa of the antral wall. Each dog underwent three recordings under different conditions: (1) basal state, (2) mild surgically-induced uncoupling, and (3) severe surgically-induced uncoupling. For each condition half-hour recordings were made. The neural network was implemented according to the Learning Vector Quantization model. This is a supervised learning model of the Kohonen Self-Organizing Maps. Majority of the recordings collected from the dogs were used for network training. Remaining recordings served as a testing tool to examine the validity of the training procedure. Approximately 90% of the dogs from the neural network training set were classified properly. However, only 31% of the dogs not included in the training process were accurately diagnosed. The poor neural-network based diagnosis of recordings that did not participate in the training process might have been caused by inappropriate representation of input data. Previous research has suggested characterizing signals according to certain features of the recorded data. This method, if employed, would reduce the noise and possibly improve the diagnostic abilities of the neural network.

Classifier PGN: Classification with High Confidence Rules

ACM Computing Classification System (1998): H.2.8, H.3.3.

Regularity analysis of nocturnal oximetry recordings to assist in the diagnosis of sleep apnoea syndrome

The relationship between sleep apnoea–hypopnoea syndrome (SAHS) severity and the regularity of nocturnal oxygen saturation (SaO2) recordings was analysed. Three different methods were proposed to quantify regularity: approximate entropy (AEn), sample entropy (SEn) and kernel entropy (KEn). A total of 240 subjects suspected of suffering from SAHS took part in the study. They were randomly divided into a training set (96 subjects) and a test set (144 subjects) for the adjustment and assessment of the proposed methods, respectively. According to the measurements provided by AEn, SEn and KEn, higher irregularity of oximetry signals is associated with SAHS-positive patients. Receiver operating characteristic (ROC) and Pearson correlation analyses showed that KEn was the most reliable predictor of SAHS. It provided an area under the ROC curve of 0.91 in two-class classification of subjects as SAHS-negative or SAHS-positive. Moreover, KEn measurements from oximetry data exhibited a linear dependence on the apnoea–hypopnoea index, as shown by a correlation coefficient of 0.87. Therefore, these measurements could be used for the development of simplified diagnostic techniques in order to reduce the demand for polysomnographies. Furthermore, KEn represents a convincing alternative to AEn and SEn for the diagnostic analysis of noisy biomedical signals.

An incremental multilinear system for human face learning and recognition

This dissertation establishes a novel system for human face learning and recognition based on incremental multilinear Principal Component Analysis (PCA). Most of the existing face recognition systems need training data during the learning process. The system as proposed in this dissertation utilizes an unsupervised or weakly supervised learning approach, in which the learning phase requires a minimal amount of training data. It also overcomes the inability of traditional systems to adapt to the testing phase as the decision process for the newly acquired images continues to rely on that same old training data set. Consequently when a new training set is to be used, the traditional approach will require that the entire eigensystem will have to be generated again. However, as a means to speed up this computational process, the proposed method uses the eigensystem generated from the old training set together with the new images to generate more effectively the new eigensystem in a so-called incremental learning process. In the empirical evaluation phase, there are two key factors that are essential in evaluating the performance of the proposed method: (1) recognition accuracy and (2) computational complexity. In order to establish the most suitable algorithm for this research, a comparative analysis of the best performing methods has been carried out first. The results of the comparative analysis advocated for the initial utilization of the multilinear PCA in our research. As for the consideration of the issue of computational complexity for the subspace update procedure, a novel incremental algorithm, which combines the traditional sequential Karhunen-Loeve (SKL) algorithm with the newly developed incremental modified fast PCA algorithm, was established. In order to utilize the multilinear PCA in the incremental process, a new unfolding method was developed to affix the newly added data at the end of the previous data. The results of the incremental process based on these two methods were obtained to bear out these new theoretical improvements. Some object tracking results using video images are also provided as another challenging task to prove the soundness of this incremental multilinear learning method.

Detecção e diagnostico de agarramento em válvulas posicionadoras

Valve stiction, or static friction, in control loops is a common problem in modern industrial processes. Recently, many studies have been developed to understand, reproduce and detect such problem, but quantification still remains a challenge. Since the valve position (mv) is normally unknown in an industrial process, the main challenge is to diagnose stiction knowing only the output signals of the process (pv) and the control signal (op). This paper presents an Artificial Neural Network approach in order to detect and quantify the amount of static friction using only the pv and op information. Different methods for preprocessing the training set of the neural network are presented. Those methods are based on the calculation of centroid and Fourier Transform. The proposal is validated using a simulated process and the results show a satisfactory measurement of stiction.

Labeling large scale social media data using budget-driven One-class SVM classification

The social media classification problems draw more and more attention in the past few years. With the rapid development of Internet and the popularity of computers, there is astronomical amount of information in the social network (social media platforms). The datasets are generally large scale and are often corrupted by noise. The presence of noise in training set has strong impact on the performance of supervised learning (classification) techniques. A budget-driven One-class SVM approach is presented in this thesis that is suitable for large scale social media data classification. Our approach is based on an existing online One-class SVM learning algorithm, referred as STOCS (Self-Tuning One-Class SVM) algorithm. To justify our choice, we first analyze the noise-resilient ability of STOCS using synthetic data. The experiments suggest that STOCS is more robust against label noise than several other existing approaches. Next, to handle big data classification problem for social media data, we introduce several budget driven features, which allow the algorithm to be trained within limited time and under limited memory requirement. Besides, the resulting algorithm can be easily adapted to changes in dynamic data with minimal computational cost. Compared with two state-of-the-art approaches, Lib-Linear and kNN, our approach is shown to be competitive with lower requirements of memory and time.

Learning from Geometry

Subspaces and manifolds are two powerful models for high dimensional signals. Subspaces model linear correlation and are a good fit to signals generated by physical systems, such as frontal images of human faces and multiple sources impinging at an antenna array. Manifolds model sources that are not linearly correlated, but where signals are determined by a small number of parameters. Examples are images of human faces under different poses or expressions, and handwritten digits with varying styles. However, there will always be some degree of model mismatch between the subspace or manifold model and the true statistics of the source. This dissertation exploits subspace and manifold models as prior information in various signal processing and machine learning tasks.

A near-low-rank Gaussian mixture model measures proximity to a union of linear or affine subspaces. This simple model can effectively capture the signal distribution when each class is near a subspace. This dissertation studies how the pairwise geometry between these subspaces affects classification performance. When model mismatch is vanishingly small, the probability of misclassification is determined by the product of the sines of the principal angles between subspaces. When the model mismatch is more significant, the probability of misclassification is determined by the sum of the squares of the sines of the principal angles. Reliability of classification is derived in terms of the distribution of signal energy across principal vectors. Larger principal angles lead to smaller classification error, motivating a linear transform that optimizes principal angles. This linear transformation, termed TRAIT, also preserves some specific features in each class, being complementary to a recently developed Low Rank Transform (LRT). Moreover, when the model mismatch is more significant, TRAIT shows superior performance compared to LRT.

The manifold model enforces a constraint on the freedom of data variation. Learning features that are robust to data variation is very important, especially when the size of the training set is small. A learning machine with large numbers of parameters, e.g., deep neural network, can well describe a very complicated data distribution. However, it is also more likely to be sensitive to small perturbations of the data, and to suffer from suffer from degraded performance when generalizing to unseen (test) data.

From the perspective of complexity of function classes, such a learning machine has a huge capacity (complexity), which tends to overfit. The manifold model provides us with a way of regularizing the learning machine, so as to reduce the generalization error, therefore mitigate overfiting. Two different overfiting-preventing approaches are proposed, one from the perspective of data variation, the other from capacity/complexity control. In the first approach, the learning machine is encouraged to make decisions that vary smoothly for data points in local neighborhoods on the manifold. In the second approach, a graph adjacency matrix is derived for the manifold, and the learned features are encouraged to be aligned with the principal components of this adjacency matrix. Experimental results on benchmark datasets are demonstrated, showing an obvious advantage of the proposed approaches when the training set is small.

Stochastic optimization makes it possible to track a slowly varying subspace underlying streaming data. By approximating local neighborhoods using affine subspaces, a slowly varying manifold can be efficiently tracked as well, even with corrupted and noisy data. The more the local neighborhoods, the better the approximation, but the higher the computational complexity. A multiscale approximation scheme is proposed, where the local approximating subspaces are organized in a tree structure. Splitting and merging of the tree nodes then allows efficient control of the number of neighbourhoods. Deviation (of each datum) from the learned model is estimated, yielding a series of statistics for anomaly detection. This framework extends the classical {\em changepoint detection} technique, which only works for one dimensional signals. Simulations and experiments highlight the robustness and efficacy of the proposed approach in detecting an abrupt change in an otherwise slowly varying low-dimensional manifold.

Chironomid-inferred July-air temperature of Lago Piccolo di Avigliana

Chironomid headcapsules were used to reconstruct late glacial and early-Holocene summer temperatures at Lago Piccolo di Avigliana (LPA). Two training sets (northern Sweden, North America) were used to infer temperatures. The reconstructed patterns of temperature change agreed well with the GRIP and NGRIP d18O records. Inferred temperatures were high during the Bølling (ca 19 °C), slowly decreased to ca 17.5 °C during the Allerød, reached lowest temperatures (ca 16 °C) during the Younger Dryas, and increased to ca. 18.5 °C during the Preboreal. The amplitudes of change at climate transitions (i.e. Oldest Dryas/Bølling: 3 °C, Allerød/Younger Dryas: 1.5 °C, and Younger Dryas/Preboreal: 2.5 °C) were smaller than in the northern Alps but similar to those recorded at another site in northeastern Italy. Our results suggest that (1) Allerød temperatures were higher in the southern Alps and (2) higher during the Preboreal (1 °C) than during the Allerød. These differences might provide an explanation for the different responses of terrestrial-vegetation to late glacial and early-Holocene climatic changes in the two regions. Other sites on both sides of the Alps should be studied to confirm these two hypotheses.