Feature selection by multi-objective optimisation: Application to network anomaly detection by hierarchical self-organising maps

Feature selection is an important and active issue in clustering and classification problems. By choosing an adequate feature subset, a dataset dimensionality reduction is allowed, thus contributing to decreasing the classification computational complexity, and to improving the classifier performance by avoiding redundant or irrelevant features. Although feature selection can be formally defined as an optimisation problem with only one objective, that is, the classification accuracy obtained by using the selected feature subset, in recent years, some multi-objective approaches to this problem have been proposed. These either select features that not only improve the classification accuracy, but also the generalisation capability in case of supervised classifiers, or counterbalance the bias toward lower or higher numbers of features that present some methods used to validate the clustering/classification in case of unsupervised classifiers. The main contribution of this paper is a multi-objective approach for feature selection and its application to an unsupervised clustering procedure based on Growing Hierarchical Self-Organising Maps (GHSOMs) that includes a new method for unit labelling and efficient determination of the winning unit. In the network anomaly detection problem here considered, this multi-objective approach makes it possible not only to differentiate between normal and anomalous traffic but also among different anomalies. The efficiency of our proposals has been evaluated by using the well-known DARPA/NSL-KDD datasets that contain extracted features and labelled attacks from around 2 million connections. The selected feature sets computed in our experiments provide detection rates up to 99.8% with normal traffic and up to 99.6% with anomalous traffic, as well as accuracy values up to 99.12%.

Semantic localization in the PCL library

The semantic localization problem in robotics consists in determining the place where a robot is located by means of semantic categories. The problem is usually addressed as a supervised classification process, where input data correspond to robot perceptions while classes to semantic categories, like kitchen or corridor. In this paper we propose a framework, implemented in the PCL library, which provides a set of valuable tools to easily develop and evaluate semantic localization systems. The implementation includes the generation of 3D global descriptors following a Bag-of-Words approach. This allows the generation of fixed-dimensionality descriptors from any type of keypoint detector and feature extractor combinations. The framework has been designed, structured and implemented to be easily extended with different keypoint detectors, feature extractors as well as classification models. The proposed framework has also been used to evaluate the performance of a set of already implemented descriptors, when used as input for a specific semantic localization system. The obtained results are discussed paying special attention to the internal parameters of the BoW descriptor generation process. Moreover, we also review the combination of some keypoint detectors with different 3D descriptor generation techniques.

Computational Methods for Identifying Hypertrophic Cardiomyopathy using 12-lead ECG

Hypertrophic cardiomyopathy (HCM) is a cardiovascular disease where the heart muscle is partially thickened and blood flow is - potentially fatally - obstructed. It is one of the leading causes of sudden cardiac death in young people. Electrocardiography (ECG) and Echocardiography (Echo) are the standard tests for identifying HCM and other cardiac abnormalities. The American Heart Association has recommended using a pre-participation questionnaire for young athletes instead of ECG or Echo tests due to considerations of cost and time involved in interpreting the results of these tests by an expert cardiologist. Initially we set out to develop a classifier for automated prediction of young athletes’ heart conditions based on the answers to the questionnaire. Classification results and further in-depth analysis using computational and statistical methods indicated significant shortcomings of the questionnaire in predicting cardiac abnormalities. Automated methods for analyzing ECG signals can help reduce cost and save time in the pre-participation screening process by detecting HCM and other cardiac abnormalities. Therefore, the main goal of this dissertation work is to identify HCM through computational analysis of 12-lead ECG. ECG signals recorded on one or two leads have been analyzed in the past for classifying individual heartbeats into different types of arrhythmia as annotated primarily in the MIT-BIH database. In contrast, we classify complete sequences of 12-lead ECGs to assign patients into two groups: HCM vs. non-HCM. The challenges and issues we address include missing ECG waves in one or more leads and the dimensionality of a large feature-set. We address these by proposing imputation and feature-selection methods. We develop heartbeat-classifiers by employing Random Forests and Support Vector Machines, and propose a method to classify full 12-lead ECGs based on the proportion of heartbeats classified as HCM. The results from our experiments show that the classifiers developed using our methods perform well in identifying HCM. Thus the two contributions of this thesis are the utilization of computational and statistical methods for discovering shortcomings in a current screening procedure and the development of methods to identify HCM through computational analysis of 12-lead ECG signals.

Essays in macroeconometrics

The knowledge of the current state of the economy is crucial for policy makers, economists and analysts. However, a key economic variable, the gross domestic product (GDP), are typically colected on a quartely basis and released with substancial delays by the national statistical agencies. The first aim of this paper is to use a dynamic factor model to forecast the current russian GDP, using a set of timely monthly information. This approach can cope with the typical data flow problems of non-synchronous releases, mixed frequency and the curse of dimensionality. Given that Russian economy is largely dependent on the commodity market, our second motivation relates to study the effects of innovations in the russian macroeconomic fundamentals on commodity price predictability. We identify these innovations through a news index which summarizes deviations of offical data releases from the expectations generated by the DFM and perform a forecasting exercise comparing the performance of different models.

Particel size distribution and microstructures of sediments from ODP Leg 205 sites (Table 1)

Ocean Drilling Program Legs 170 and 205 offshore Costa Rica provide structural observations which support a new model for the geometry and deformation response to the seismic cycle of the frontal sedimentary prism and decollement. The model is based on drillcore, thin section, and electron microscope observations. The decollement damage zone is a few tens of meters in width, it develops mainly within the frontal prism. A clear cm-thick fault core is observed 1.6 km from the trench. The lower boundary of the fault core is coincident with the lithological boundary between the frontal prism and the hemipelagic and pelagic sediment of the Cocos plate. Breccia clast distributions in the upper portion of the decollement damage zone were studied through fractal analysis. This analysis shows that the fractal dimension changes with brecciated fragment size, implying that deformation was not accommodated by self-similar fracturing. A higher fractal dimensionality correlates with smaller particle size, which indicates that different or additional grain-size reduction processes operated during shearing. The co-existence of two distinct fracturing processes is also confirmed by microscopic analysis in which extension fracturing in the upper part of the damage zone farthest from the fault core is frequent, while both extension and shear fracturing occur approaching the fault core. The coexistence of extensional and shear fracturing seems to be best explained by fluid pressure variations in response to variations of the compressional regime during the seismic cycle. During the co-seismic event, sub-horizontal compression and fluid pressure increase, triggering shear fracturing and fluid expulsion. Fractures migrate upward with fluids, contributing to the asymmetric shape of the decollement, while slip propagates. In the inter-seismic interval the frontal prismrelaxes and fluid pressure drops. The frontal prismgoes into diffuse extension during the intervalwhen plate convergence is accommodated by creep along the ductile fault core. The fault core is typically a barrier to deformation, which is explained by its weak, but impermeable, nature. The localized development of a damage zone beneath the fault core is characterized by shear fracturing that appears as the result of local strengthening of the detachment.

Demographic and Geographic Heterogeneity between Subgroups of Cyclists

Thesis (Master's)--University of Washington, 2016-06

What is the Impact of the Underserved Pathway program on Entering an Underserved Family Medicine Residency? A Comparison of Three Approaches for Estimating the Average Treatment Effect

Thesis (Master's)--University of Washington, 2016-06

Functionality, Robustness and Control of Nonlinear Network Dynamics: Modeling and Understanding the C. elegans Connectome

Thesis (Ph.D.)--University of Washington, 2016-06

Energy level statistics for models of coupled single-mode Bose-Einstein condensates

We study the distribution of energy level spacings in two models describing coupled single-mode Bose-Einstein condensates. Both models have a fixed number of degrees of freedom, which is small compared to the number of interaction parameters, and is independent of the dimensionality of the Hilbert space. We find that the distribution follows a universal Poisson form independent of the choice of coupling parameters, which is indicative of the integrability of both models. These results complement those for integrable lattice models where the number of degrees of freedom increases with increasing dimensionality of the Hilbert space. Finally, we also show that for one model the inclusion of an additional interaction which breaks the integrability leads to a non-Poisson distribution.

Fock-state dynamics in Raman photoassociation of Bose-Einstein condensates

By stochastic modeling of the process of Raman photoassociation of Bose-Einstein condensates, we show that, the farther the initial quantum state is from a coherent state, the farther the one-dimensional predictions are from those of the commonly used zero-dimensional approach. We compare the dynamics of condensates, initially in different quantum states, finding that, even when the quantum prediction for an initial coherent state is relatively close to the Gross-Pitaevskii prediction, an initial Fock state gives qualitatively different predictions. We also show that this difference is not present in a single-mode type of model, but that the quantum statistics assume a more important role as the dimensionality of the model is increased. This contrasting behavior in different dimensions, well known with critical phenomena in statistical mechanics, makes itself plainly visible here in a mesoscopic system and is a strong demonstration of the need to consider physically realistic models of interacting condensates.

The positive impact of multiple sclerosis (MS) on carers: Associations between carer benefit finding and positive and negative adjustment domains

Purpose. This study examined benefit finding in MS carers including the dimensionality of benefit finding, relations between carer and care recipient benefit finding, and the effects of carer benefit finding on carer positive and negative adjustment domains. Method. A total of 267 carers and their care recipients completed questionnaires at Time 1 and 3 months later, Time 2 (n=155). Illness data were collected at Time 1, and number of problems, stress appraisal, benefit finding, negative (global distress, negative affect) and positive (life satisfaction, positive affect, dyadic adjustment) adjustment domains were measured at Time 2. Results. Qualitative data revealed seven benefit finding themes, two of which were adequately represented by the Benefit Finding Scale (BFS) [1] (Mohr et al. Health Psychology 1999; 18: 376). Factor analyses indicated two factors (Personal Growth, Family Relations Growth) which were psychometrically sound and showed differential relations with illness and adjustment domains. Although care recipients reported higher levels of benefit finding than carers, their benefit finding reports regarding personal growth were correlated. The carer BFS factors were positively related to carer and care recipient dyadic adjustment. Care recipient benefit finding was unrelated to carer adjustment domains. After controlling for the effects of demographics, care recipient characteristics, problems and appraisal, carer benefit finding was related to carer positive adjustment domains and unrelated to carer negative adjustment domains. Conclusion. Findings support the role of benefit finding in sustaining positive psychological states and the communal search for meaning within carer-care recipient dyads.

A consumer-based method for retailer equity measurement: Results of an empirical study

This research extends the consumer-based brand equity measurement approach to the measurement of the equity associated with retailers. This paper also addresses some of the limitations associated with current retailer equity measurement such as a lack of clarity regarding its nature and dimensionality. We conceptualise retailer equity as a four-dimensional construct comprising retailer awareness, retailer associations, perceived retailer quality, and retailer loyalty. The paper reports the result of an empirical study of a convenience sample of 601 shopping mall consumers at an Australian state capital city. Following a confirmatory factor analysis using structural equation modelling to examine the dimensionality of the retailer equity construct, the proposed model is tested for two retailer categories: department stores and speciality stores. Results confirm the hypothesised four-dimensional structure.

Hierarchical indexing structure for efficient similarity search in video retrieval

With the rapid increase in both centralized video archives and distributed WWW video resources, content-based video retrieval is gaining its importance. To support such applications efficiently, content-based video indexing must be addressed. Typically, each video is represented by a sequence of frames. Due to the high dimensionality of frame representation and the large number of frames, video indexing introduces an additional degree of complexity. In this paper, we address the problem of content-based video indexing and propose an efficient solution, called the Ordered VA-File (OVA-File) based on the VA-file. OVA-File is a hierarchical structure and has two novel features: 1) partitioning the whole file into slices such that only a small number of slices are accessed and checked during k Nearest Neighbor (kNN) search and 2) efficient handling of insertions of new vectors into the OVA-File, such that the average distance between the new vectors and those approximations near that position is minimized. To facilitate a search, we present an efficient approximate kNN algorithm named Ordered VA-LOW (OVA-LOW) based on the proposed OVA-File. OVA-LOW first chooses possible OVA-Slices by ranking the distances between their corresponding centers and the query vector, and then visits all approximations in the selected OVA-Slices to work out approximate kNN. The number of possible OVA-Slices is controlled by a user-defined parameter delta. By adjusting delta, OVA-LOW provides a trade-off between the query cost and the result quality. Query by video clip consisting of multiple frames is also discussed. Extensive experimental studies using real video data sets were conducted and the results showed that our methods can yield a significant speed-up over an existing VA-file-based method and iDistance with high query result quality. Furthermore, by incorporating temporal correlation of video content, our methods achieved much more efficient performance.

Anomalous Excitation Spectra of Frustrated Quantum Antiferromagnets

We use series expansions to study the excitation spectra of spin-1/2 antiferromagnets on anisotropic triangular lattices. For the isotropic triangular lattice model (TLM), the high-energy spectra show several anomalous features that differ strongly from linear spin-wave theory (LSWT). Even in the Neel phase, the deviations from LSWT increase sharply with frustration, leading to rotonlike minima at special wave vectors. We argue that these results can be interpreted naturally in a spinon language and provide an explanation for the previously observed anomalous finite-temperature properties of the TLM. In the coupled-chains limit, quantum renormalizations strongly enhance the one-dimensionality of the spectra, in agreement with experiments on Cs2CuCl4.

Efficient accommodation of local minima in watershed model calibration

The Gauss-Marquardt-Levenberg (GML) method of computer-based parameter estimation, in common with other gradient-based approaches, suffers from the drawback that it may become trapped in local objective function minima, and thus report optimized parameter values that are not, in fact, optimized at all. This can seriously degrade its utility in the calibration of watershed models where local optima abound. Nevertheless, the method also has advantages, chief among these being its model-run efficiency, and its ability to report useful information on parameter sensitivities and covariances as a by-product of its use. It is also easily adapted to maintain this efficiency in the face of potential numerical problems (that adversely affect all parameter estimation methodologies) caused by parameter insensitivity and/or parameter correlation. The present paper presents two algorithmic enhancements to the GML method that retain its strengths, but which overcome its weaknesses in the face of local optima. Using the first of these methods an intelligent search for better parameter sets is conducted in parameter subspaces of decreasing dimensionality when progress of the parameter estimation process is slowed either by numerical instability incurred through problem ill-posedness, or when a local objective function minimum is encountered. The second methodology minimizes the chance of successive GML parameter estimation runs finding the same objective function minimum by starting successive runs at points that are maximally removed from previous parameter trajectories. As well as enhancing the ability of a GML-based method to find the global objective function minimum, the latter technique can also be used to find the locations of many non-global optima (should they exist) in parameter space. This can provide a useful means of inquiring into the well-posedness of a parameter estimation problem, and for detecting the presence of bimodal parameter and predictive probability distributions. The new methodologies are demonstrated by calibrating a Hydrological Simulation Program-FORTRAN (HSPF) model against a time series of daily flows. Comparison with the SCE-UA method in this calibration context demonstrates a high level of comparative model run efficiency for the new method. (c) 2006 Elsevier B.V. All rights reserved.