An intelligent configuration-sampling based local motion planner for robotic manipulators

Manipulator motion planning is a task which relies heavily on the construction of a configuration space prior to path planning. However when fast real-time motion is needed, the full construction of the manipulator's high-dimensional configu-ration space can be too slow and expensive. Alternative planning methods, which avoid this full construction of the manipulator's configuration space are needed to solve this problem. Here, one such existing local planning method for manipulators based on configuration-sampling and subgoal-selection has been extended. Using a modified Artificial Potential Fields (APF) function, goal-configuration sampling and a novel subgoal selection method, it provides faster, more optimal paths than the previously proposed work. Simulation results show a decrease in both runtime and path lengths, along with a decrease in unexpected local minimum and crashing issues.

The effect of probe interval estimation on attack detection performance of a WLAN independent intrusion detection system

A new niche of densely populated, unprotected networks is becoming more prevalent in public areas such as Shopping Malls, defined here as independent open-access networks, which have attributes that make attack detection more challenging than in typical enterprise networks. To address these challenges, new detection systems which do not rely on knowledge of internal device state are investigated here. This paper shows that this lack of state information requires an additional metric (The exchange timeout window) for detection of WLAN Denial of Service Probe Flood attacks. Variability in this metric has a significant influence on the ability of a detection system to reliably detect the presence of attacks. A parameter selection method is proposed which is shown to provide reliability and repeatability in attack detection in WLANs. Results obtained from ongoing live trials are presented that demonstrate the importance of accurately estimating probe request and probe response timeouts in future Independent Intrusion Detection Systems.

The prognostic and therapeutic value of EpHA2 in early colorectal cancer (CRC).

Background: EpHA2 is a 130 kD transmembrane glycoprotein belonging to ephrin receptor subfamily and involved in angiogenesis/tumour neovascularisation. High EpHA2 mRNA level has recently been implicated in cetuximab resistance. Previously, we found high EpHA2 levels in a panel of invasive colorectal cancer (CRC) cells, which was associated with high levels of stem-cell marker CD44. Our aim was to investigate the prognostic value of EpHA2 and subsequently correlate expression levels to known clinico-pathological variables in early stage CRC. Methods: Tissue samples from 509 CRC patients were analysed. EpHA2 expression was measured using IHC. Kaplan-Meier graphs were used. Univariate and multivariate analyses employed Cox Proportional Hazards Ratio (HR) method. A backward selection method (Akaike’s information criterion) was used to determine a refined multivariate model. Results: EpHA2 was highly expressed in CRC adenocarcinoma compared to matched normal colon tissue. In support of our preclinical invasive models, strong correlation was found between EpHA2 expression and CD44 and Lgr5 staining (p<0.001). In addition, high EpHA2 expression significantly correlated with vascular invasion (p=0.03).HR for OS for stage II/III patients with high EpHA2 expression was 1.69 (95%CI: 1.164-2.439; p=0.003). When stage II/III was broken down into individual stages, there was significant correlation between high EpHA2 expression and poor 5-years OS in stage II patients (HR: 2.18; 95%CI: 1.28-3.71; p=0.005).HR in the stage III group showed a trend to statistical significance (HR: 1.48; 95%CI=0.87-2.51; p=0.05). In both univariate and multivariate analyses of stage II patients, high EpHA2 expression was the only significant factor and was retained in the final multivariate model. Higher levels of EpHA2 were noted in our RAS and BRAF mutant CRC cells, and silencing EpHA2 resulted in significant decreases in migration/invasion in parental and invasive CRC sublines. Correlation between KRAS/NRAS/BRAFmutational status and EpHA2 expression in clinical samples is ongoing. Conclusions: Taken together, our study is the first to indicate that EpHA2 expression is a predictor of poor clinical outcome and a potential novel target in early stage CRC.

Pregnancy and newborns disorders followed by urine metabolomics

Chapter 1 introduces the scope of the work by identifying the clinically relevant prenatal disorders and presently available diagnostic methods. The methodology followed in this work is presented, along with a brief account of the principles of the analytical and statistical tools employed. A thorough description of the state of the art of metabolomics in prenatal research concludes the chapter, highlighting the merit of this novel strategy to identify robust disease biomarkers. The scarce use of maternal and newborn urine in previous reports enlightens the relevance of this work. Chapter 2 presents a description of all the experimental details involved in the work performed, comprising sampling, sample collection and preparation issues, data acquisition protocols and data analysis procedures. The proton Nuclear Magnetic Resonance (NMR) characterization of maternal urine composition in healthy pregnancies is presented in Chapter 3. The urinary metabolic profile characteristic of each pregnancy trimester was defined and a 21-metabolite signature found descriptive of the metabolic adaptations occurring throughout pregnancy. 8 metabolites were found, for the first time to our knowledge, to vary in connection to pregnancy, while known metabolic effects were confirmed. This chapter includes a study of the effects of non-fasting (used in this work) as a possible confounder. Chapter 4 describes the metabolomic study of 2nd trimester maternal urine for the diagnosis of fetal disorders and prediction of later-developing complications. This was achieved by applying a novel variable selection method developed in the context of this work. It was found that fetal malformations (FM) (and, specifically those of the central nervous system, CNS) and chromosomal disorders (CD) (and, specifically, trisomy 21, T21) are accompanied by changes in energy, amino acids, lipids and nucleotides metabolic pathways, with CD causing a further deregulation in sugars metabolism, urea cycle and/or creatinine biosynthesis. Multivariate analysis models´ validation revealed classification rates (CR) of 84% for FM (87%, CNS) and 85% for CD (94%, T21). For later-diagnosed preterm delivery (PTD), preeclampsia (PE) and intrauterine growth restriction (IUGR), it is found that urinary NMR profiles have early predictive value, with CRs ranging from 84% for PTD (11-20 gestational weeks, g.w., prior to diagnosis), 94% for PE (18-24 g.w. pre-diagnosis) and 94% for IUGR (2-22 g.w. pre-diagnosis). This chapter includes results obtained for an ultraperformance liquid chromatography-mass spectrometry (UPLC-MS) study of pre-PTD samples and correlation with NMR data. One possible marker was detected, although its identification was not possible. Chapter 5 relates to the NMR metabolomic study of gestational diabetes mellitus (GDM), establishing a potentially predictive urinary metabolic profile for GDM, 2-21 g.w. prior to diagnosis (CR 83%). Furthermore, the NMR spectrum was shown to carry information on individual phenotypes, able to predict future insulin treatment requirement (CR 94%). Chapter 6 describes results that demonstrate the impact of delivery mode (CR 88%) and gender (CR 76%) on newborn urinary profile. It was also found that newborn prematurity, respiratory depression, large for gestational age growth and malformations induce relevant metabolic perturbations (CR 82-92%), as well as maternal conditions, namely GDM (CR 82%) and maternal psychiatric disorders (CR 91%). Finally, the main conclusions of this thesis are presented in Chapter 7, highlighting the value of maternal or newborn urine metabolomics for pregnancy monitoring and disease prediction, towards the development of new early and non-invasive diagnostic methods.

Assessing the effects of pseudo-absences on predictive distribution model performance

Modelling species distributions with presence data from atlases, museum collections and databases is challenging. In this paper, we compare seven procedures to generate pseudoabsence data, which in turn are used to generate GLM-logistic regressed models when reliable absence data are not available. We use pseudo-absences selected randomly or by means of presence-only methods (ENFA and MDE) to model the distribution of a threatened endemic Iberian moth species (Graellsia isabelae). The results show that the pseudo-absence selection method greatly influences the percentage of explained variability, the scores of the accuracy measures and, most importantly, the degree of constraint in the distribution estimated. As we extract pseudo-absences from environmental regions further from the optimum established by presence data, the models generated obtain better accuracy scores, and over-prediction increases. When variables other than environmental ones influence the distribution of the species (i.e., non-equilibrium state) and precise information on absences is non-existent, the random selection of pseudo-absences or their selection from environmental localities similar to those of species presence data generates the most constrained predictive distribution maps, because pseudo-absences can be located within environmentally suitable areas. This study showsthat ifwe do not have reliable absence data, the method of pseudo-absence selection strongly conditions the obtained model, generating different model predictions in the gradient between potential and realized distributions.

Detecting Adverse Events in Clinical Trial Free Text

Thesis (Master's)--University of Washington, 2013

Statistical classification of road pavements using near field vehicle rolling noise measurements

Low noise surfaces have been increasingly considered as a viable and cost-effective alternative to acoustical barriers. However, road planners and administrators frequently lack information on the correlation between the type of road surface and the resulting noise emission profile. To address this problem, a method to identify and classify different types of road pavements was developed, whereby near field road noise is analyzed using statistical learning methods. The vehicle rolling sound signal near the tires and close to the road surface was acquired by two microphones in a special arrangement which implements the Close-Proximity method. A set of features, characterizing the properties of the road pavement, was extracted from the corresponding sound profiles. A feature selection method was used to automatically select those that are most relevant in predicting the type of pavement, while reducing the computational cost. A set of different types of road pavement segments were tested and the performance of the classifier was evaluated. Results of pavement classification performed during a road journey are presented on a map, together with geographical data. This procedure leads to a considerable improvement in the quality of road pavement noise data, thereby increasing the accuracy of road traffic noise prediction models.

Energy and delay trade-off of the GTS allocation mechanism in IEEE 802.15.4 for wireless sensor networks

The IEEE 802.15.4 protocol proposes a flexible communication solution for Low-Rate Wireless Personal Area Networks (LR-WPAN) including wireless sensor networks (WSNs). It presents the advantage to fit different requirements of potential applications by adequately setting its parameters. When in beaconenabled mode, the protocol can provide timeliness guarantees by using its Guaranteed Time Slot (GTS) mechanism. However, power-efficiency and timeliness guarantees are often two antagonistic requirements in wireless sensor networks. The purpose of this paper is to analyze and propose a methodology for setting the relevant parameters of IEEE 802.15.4-compliant WSNs that takes into account a proper trade-off between power-efficiency and delay bound guarantees. First, we propose two accurate models of service curves for a GTS allocation as a function of the IEEE 802.15.4 parameters, using Network Calculus formalism. We then evaluate the delay bound guaranteed by a GTS allocation and express it as a function of the duty cycle. Based on the relation between the delay requirement and the duty cycle, we propose a power-efficient superframe selection method that simultaneously reduces power consumption and enables meeting the delay requirements of real-time flows allocating GTSs. The results of this work may pave the way for a powerefficient management of the GTS mechanism in an IEEE 802.15.4 cluster.

Gait Analysis for Classification

This thesis describes a representation of gait appearance for the purpose of person identification and classification. This gait representation is based on simple localized image features such as moments extracted from orthogonal view video silhouettes of human walking motion. A suite of time-integration methods, spanning a range of coarseness of time aggregation and modeling of feature distributions, are applied to these image features to create a suite of gait sequence representations. Despite their simplicity, the resulting feature vectors contain enough information to perform well on human identification and gender classification tasks. We demonstrate the accuracy of recognition on gait video sequences collected over different days and times and under varying lighting environments. Each of the integration methods are investigated for their advantages and disadvantages. An improved gait representation is built based on our experiences with the initial set of gait representations. In addition, we show gender classification results using our gait appearance features, the effect of our heuristic feature selection method, and the significance of individual features.

Sparse kernel modelling: a unified approach

A unified approach is proposed for sparse kernel data modelling that includes regression and classification as well as probability density function estimation. The orthogonal-least-squares forward selection method based on the leave-one-out test criteria is presented within this unified data-modelling framework to construct sparse kernel models that generalise well. Examples from regression, classification and density estimation applications are used to illustrate the effectiveness of this generic sparse kernel data modelling approach.

Integrating features for man-made target tracking from FLIR image sequence using particle filter

A new man-made target tracking algorithm integrating features from (Forward Looking InfraRed) image sequence is presented based on particle filter. Firstly, a multiscale fractal feature is used to enhance targets in FLIR images. Secondly, the gray space feature is defined by Bhattacharyya distance between intensity histograms of the reference target and a sample target from MFF (Multi-scale Fractal Feature) image. Thirdly, the motion feature is obtained by differencing between two MFF images. Fourthly, a fusion coefficient can be automatically obtained by online feature selection method for features integrating based on fuzzy logic. Finally, a particle filtering framework is developed to fulfill the target tracking. Experimental results have shown that the proposed algorithm can accurately track weak or small man-made target in FLIR images with complicated background. The algorithm is effective, robust and satisfied to real time tracking.

A comparative review of dimension reduction methods in approximate Bayesian computation

Approximate Bayesian computation (ABC) methods make use of comparisons between simulated and observed summary statistics to overcome the problem of computationally intractable likelihood functions. As the practical implementation of ABC requires computations based on vectors of summary statistics, rather than full data sets, a central question is how to derive low-dimensional summary statistics from the observed data with minimal loss of information. In this article we provide a comprehensive review and comparison of the performance of the principal methods of dimension reduction proposed in the ABC literature. The methods are split into three nonmutually exclusive classes consisting of best subset selection methods, projection techniques and regularization. In addition, we introduce two new methods of dimension reduction. The first is a best subset selection method based on Akaike and Bayesian information criteria, and the second uses ridge regression as a regularization procedure. We illustrate the performance of these dimension reduction techniques through the analysis of three challenging models and data sets.

Prediction error property of the Lasso Estimator and its generalization

The lasso procedure is an estimator-shrinkage and variable selection method. This paper shows that there always exists an interval of tuning parameter values such that the corresponding mean squared prediction error for the lasso estimator is smaller than for the ordinary least squares estimator. For an estimator satisfying some condition such as unbiasedness, the paper defines the corresponding generalized lasso estimator. Its mean squared prediction error is shown to be smaller than that of the estimator for values of the tuning parameter in some interval. This implies that all unbiased estimators are not admissible. Simulation results for five models support the theoretical results.

Spam filtering using classification algorithms

This thesis proposes an innovative adaptive multi-classifier spam filtering model, with a grey-list analyser and a dynamic feature selection method, to overcome false-positive problems in email classification. It also presents additional techniques to minimize the added complexity. Empirical evidence indicates the success of this model over existing approaches.

Replacement of RNA hairpins by in vitro selected tetranucleotides

An in vitro selection method based on the autolytic cleavage of yeast tRNAPhe by Pb2+ was applied to obtain tRNA derivatives with the anticodon hairpin replaced by four single-stranded nucleotides. Based on the rates of the site-specific cleavage by Pb2+ and the presence of a specific UV-induced crosslink, certain tetranucieotide sequences allow proper folding of the rest of the tRNA molecule, wheras others do not. One such successful tetramer sequence was also used to replace the acceptor stem of yeast tRNAPhe and the anticodon hairpin of E.coli tRNAPhe without disrupting folding. These experiments suggest that certain tetramers may be able to replace structurally non essential hairpins in any RNA.