Modeling covariance structure is important for efficient estimation in longitudinal data models. Modified Cholesky decomposition (Pourahmadi, 1999) is used as an unconstrained reparameterization of the covariance matrix. The resulting new parameters have transparent statistical interpretations and are easily modeled using covariates. However, this approach is not directly applicable when the longitudinal data are unbalanced, because a Cholesky factorization for observed data that is coherent across all subjects usually does not exist. We overcome this difficulty by treating the problem as a missing data problem and employing a generalized EM algorithm to compute the ML estimators. We study the covariance matrices in both fixed-effects models and mixed-effects models for unbalanced longitudinal data. We illustrate our method by reanalyzing Kenwards (1987) cattle data and conducting simulation studies.

Usually, a large steam turbine-generator unit has itself a blade vibration mode that is close to its double electrical frequency. This mode of vibration will easily be excited by electrical load unbalance, thereby the turbine blades will be affected by this kind of vibrations, especially for the last three rows of blades. In fact, turbine generators operate in corrosive environment and undergo the statistical stress impact due to the randomly unbalanced currents. In this paper, the blades are subjected to corrosion fatigue, thereby small stress still may cause damage significantly. On the other hand, the damage caused by system unbalance is so small that people could neglect it usually. Nevertheless, for the long-term operation with lasting system unbalance, its influence on reliability may no longer be omitted. According to the gamma distribution in unbalanced negative phase current (I2), the probability level of fatigue life, the reliability against fatigue failure and crack growth of turbine blades are evaluated for three turbine-generator system in the paper. The blades with various materials, safety factors and stress concentration factors are considered in the simulations. The influence of extreme value distribution of I2 to the reliability is also investigated. According to the results, we have reason to believe that corrosion fatigue is one of causes that led to crack initiation or damage of blades under normal operation.

Anthracnose of lentil caused by the fungal pathogen Colletotrichum lentis is an economically important disease in Western Canada. The pathogen population is divided into two races (0 and 1) and two sexual incompatibility groups (IG-1 and IG-2). Resistance to anthracnose race 1 is found in cultivated Lens cultivars whereas for the more aggressive race 0, higher levels of resistance have been reported only from wild Lens species. Furthermore, C. lentis seems to only possess one (MAT1-2) of the two mating type idiomorphs commonly present in heterothallic ascomycete fungi with the typical bipolar mating system. The purpose of this study was to verify the phylogenetic relationship between race 0 and 1 isolates of C. lentis and to sequence and characterize the MAT1-2 of C. lentis. A morphological, multi-locus phylogenetic and host-range study was conducted with isolates of C. lentis, C. truncatum (from various host species and the epitype), C. destructivum, C. dematium, C. higginsianum, C. linicola and C. lindemuthianum. Sequence data from six conserved loci displayed 100% identity for C. lentis isolates of both races that formed a single cluster separate from other Colletotrichum species including C. destructivum, the epitype of C. truncatum and isolates from other hosts identified as C. truncatum. Conidia of C. lentis were slightly falcate with obtuse apices compared to cylindrical conidia with rounded ends of C. destructivum, and longer lunate to falcate conidia of the epitype C. truncatum. Host range tests undertaken on Lens culinaris, Pisum sativum, Cicer arietinum, Vicia faba, Glycine max, Phaseolus vulgaris, Phaseolus lunatus, Trifolium pratense, Medicago sativa, Medicago truncatula, Brassica chinensis and Arabidopsis thaliana under controlled environmental conditions revealed that the host ranges of C. linicola and C. higginsianum overlapped with that of lentil isolates. In contrast, the epitype specimen of C. truncatum was pathogenic on Pisum sativum, Phaseolus vulgaris, T. pratense and Medicago sativa, but not on L. culinaris. All Colletotrichum spp. infected Medicago truncatula and all but the lentil isolates caused disease on G. max. The mating type gene MAT1-2 of C. lentis contained two introns and three exons and an open reading frame of 726 bp coding for a putative protein of 241 amino acids including the high mobility group (HMG) domain characteristic of the MAT1-2 in fungi. The MAT1-2 nucleotide sequences of C. lentis isolates were identical irrespective of IG. An isolate from each of the two IGs, CT-21 (IG-2), CT-30 (IG-1) and a co-culture of CT-21 and CT-30 was used to study the expression levels of MAT1-2 at seven different in vitro time points (0h, 6h, 12h, 18h, 24h, 36, 48h after inoculation in glucose yeast media) and investigate for possible alternative splicing events. MAT1-2 expression for CT-21, CT-30 and the co-culture was observed at all seven time points indicating that it is constitutively expressed, and no differences in the transcript size were seen, ruling out the possibility of a splicing event.

The International Normalizing Ratio (INR) measures the speed at which blood clots. Healthy people have an INR of about one. Some people are at greater risk of blood clots and their physician prescribes a target INR range, generally 2-3. The farther a patient is above or below their prescribed range, the more dangerous their situation. A variety of point-of-care (POC) devices has been developed to monitor patients. The purpose of this research was to develop innovative graphics to help describe a highly unbalanced dataset and to carry out Bayesian analyses to determine which of five devices best manages patients. An initial Bayesian analysis compared a machine-identical beta-binomial model to a machine-specific beta-binomial model. The response variable was number of in-range visits. A second Bayesian analysis compared a machine-identical lognormal model, a machine-specific lognormal model, and a machine-specific lognormal model with lower therapeutic bound as a predictor. The response variable was INR. Machines were compared using posterior predictive distributions of the absolute distance outside a patient's therapeutic range. For the beta-binomial models, the machine-identical model had the lower DIC, meaning that POC device was not a strong predictor of success in keeping a patient in-range. The machine-specific lognormal model with a term for lower therapeutic bound had the lowest DIC of the three lognormal models, implying that the additional information of distance out of range revealed differences among the POC devices. Three of the machines had more favorable out-of-range distributions than the other two. Both Bayesian analyses provided useful information for medical practice in managing INR.

This thesis is an extension of recent research into the relationships between non-homothetic preferences and patterns of trade. The analysis focuses on the observed shift in consumption towards income-elastic services and, relative to agricultural goods, income-elastic manufactures associated with rising per capita incomes. In turn, the conjecture that we should witness a shift in global production and consequently a shift in trade away from primaries towards manufactured goods as the global economy develops is explored. This hypothesized change in the sectoral composition of global trade implies a change in individual country trade patterns. Specifically, the notion that a countrys exports must respond to a changing global market may help to clarify one of the principle causes of the shift towards manufacturing production among most small, trading economies.

This thesis is an extension of recent research into the relationships between non-homothetic preferences and patterns of trade. The analysis focuses on the observed shift in consumption towards income-elastic services and, relative to agricultural goods, income-elastic manufactures associated with rising per capita incomes. In turn, the conjecture that we should witness a shift in global production and consequently a shift in trade away from primaries towards manufactured goods as the global economy develops is explored. This hypothesized change in the sectoral composition of global trade implies a change in individual country trade patterns. Specifically, the notion that a countrys exports must respond to a changing global market may help to clarify one of the principle causes of the shift towards manufacturing production among most small, trading economies.

With the rapid development of power systems, more and more smart devices are installed in power industries, and each of them is gathering tons of information every day.Due to the data explosion and the difficulty of processing these data, data visualization, a big data technology, has become a trend. With the help of information technology, the visualization of real-time data has been achieved in power industries and there are multiple successful examples such as one-line diagram, load flow dashboard and equipment dashboard.In fault analysis group, digital fault recorders are essential to record and report an event.They are triggered when a fault occurs and corresponding report is generated instantly.However, people seldom utilize the historical data from DFRs to analyze the power quality issues.Therefore, this thesis presents the development of a power quality dashboard by using the collected data from DFRs.Three related power quality analyses have been accomplished in this paper: voltage and current variation, harmonics and unbalance components.Recursive algorithm is applied to compute the phasors and errors; Discrete Fourier Transform is utilized to extract harmonics from the samples; and the symmetric components are calculated by "A"-matrix transformation.The start page for the dashboard is a google map with all the DFR markers, and after double-clicking the marker, the report page will be opened.With the reports, engineers can not only monitor the event but also analyze out the possible causes and characteristics for a fault.For those renewable energy substation, the harmonic contents can be supervised so that the damages and losses can be significantly reduced by identifying the high harmonics.Ultimately, the goal of the dashboard is to achieve warning status and harmonic gradient mapping in the future. / Master of Science

In any health-care system, early identification of individuals who are most at risk of developing an illness is vital, not only to ensure that a patient is provided with the appropriate treatment, but also to avoid the considerable costs associated with unnecessary hospitalization. To achieve this goal there is a need for a breakthrough prediction method that is capable of dealing with a real world medical data which is inherently complex. In this study, we show how standard classification algorithms can be employed collectively to predict the length of stay in a hospital of a patient in the upcoming year, based on their medical history. Multiple classifiers are used to perform the prediction task, since real world medical data is significantly complex making the classification task very challenging. The data is voluminous, consists of wide range of class values some of which with a few instances, and it is highly unbalanced making the classification of minority classes very difficult. We propose two Sequential Ensemble Classification (SEC) schemes, one based on an ensemble of homogeneous classifiers, and a second based on a heterogeneous ensemble of classifiers, in three hierarchical granularity levels. The goal of using this system is to provide increased performance over the standard classifiers. This method is highly beneficial when dealing with complex data which is multi-class and highly unbalanced.

No description available.

Abstract Traditional synchronous motion control of a multi-cylinder system was always achieved by using hydraulic loops design and constrained linkage mechanisms. Therefore, these control methodologies always have many disadvantages, such as inaccuracy, cost expensive, and huge volume of the equipment, and so on. In this paper, the nonlinear control strategy was proposed to control the proportional directional valves of two-cylinder electro-hydraulic system in order to achieve synchronous motion under the consideration of unbalanced and uncertainty loading. Besides, in order to explore influence of different loading to the system, two-cylinder mechanism was designed to have individual loading device without any hardware constrain between two pistons. And the maximum loading capacity for one piston is 210kg. Due to the highly complicated coupling effect of internal pressure and flow rate for two cylinders, in this paper, feedforward controller with three fuzzy controllers was designated to overcome the problem of synchronous motion. In the first, the feedforward controller of each cylinder is developed to track a desire velocity trajectory. Then, the fuzzy control of each cylinder was specified to improve the individual tracking performance. Finally, the third fuzzy controller was performed to compensate the coupling effect of two-cylinder in order to progressively improve the performance of synchronous motion. According to the experimental results, the proposed control strategy for synchronous motion of two-cylinder system was verified and the maximum synchronous error of the total system was controlled to be within 10mm. Keyword: synchronous motion; proportional directional valve; fuzzy controller; feedforward controller