Background: Cellulitis is a common infectious disease. Although blood culture is frequently used in the diagnosis and subsequent treatment of cellulitis, it is a contentious diagnostic test. To help clinicians determine which patients should undergo blood culture for the management of cellulitis, a diagnostic scoring system referred to as the Bacteremia Score of Cellulitis was developed. Methods: Univariable and multivariable logistic regression analyses were performed as part of a retrospective cohort study of all adults diagnosed with cellulitis in a tertiary teaching hospital in Taiwan in 2013. Patients who underwent blood culture were used to develop a diagnostic prediction model where the main outcome measures were true bacteremia in cellulitis cases. Area under the receiver operating characteristics curve (AUC) was used to demonstrate the predictive power of the model, and bootstrapping was then used to validate the performance. Results: Three hundred fifty one cases with cellulitis who underwent blood culture were enrolled. The overall prevalence of true bacteremia was 33/351 cases (9.4 %). Multivariable logistic regression analysis showed optimal diagnostic discrimination for the combination of age >= 65 years (odds ratio [OR] = 3.9; 95 % confidence interval (CI), 1.5-10.1), involvement of non-lower extremities (OR = 4.0; 95 % CI, 1.5-10.6), liver cirrhosis (OR = 6.8; 95 % CI, 1.8-25.3), and systemic inflammatory response syndrome (SIRS) (OR = 15.2; 95 % CI, 4.8-48.0). These four independent factors were included in the initial formula, and the AUC for this combination of factors was 0.867 (95 % CI, 0.806-0.928). The rounded formula was 1 x (age >= 65 years) + 1.5 x (involvement of non-lower extremities) + 2 x (liver cirrhosis) + 2.5 x (SIRS). The overall prevalence of true bacteremia (9.4 %) in this study could be lowered to 1.0 % (low risk group, score <= 1.5) or raised to 14.7 % (medium risk group, score 2-3.5) and 41.2 % (high risk group, score >= 4.0), depending on different clinical scores. Conclusions: Determining the risk of bacteremia in patients with cellulitis will allow a more efficient use of blood cultures in the diagnosis and treatment of this condition. External validation of this preliminary scoring system in future trials is needed to optimize the test.

This thesis proposes an improvement to present near-shore wave prediction models. Using weakly dispersive Boussinesq theory, the shoaling of directionally spread surface gravity waves over a beach with gentle gradients in the cross-shore and alongshore directions is examined. Following Herbers and Burton (1997), the governing fluid flow equations are expanded to third order and depth-integrated over the water column. A resulting amplitude evolution equation for a spectrum of waves is derived, which is the main result of this paper. New terms in the higher order result include effects due to alongshore bottom slope, higher order cross-shore depth variations, and non-linear quartet interactions. The linear terms in this equation are verified by analytical methods using linear finite depth theory. Example computations for a monochromatic wave train over a plane beach quantify some of the improvements of this result over the lower order model. Opportunities for further development and verification of this result are proposed, and recommendations for application of the result in its present form are outlined. / US Navy (USN) author

The problem with which this study is concerned is the development of a mathematical model and theory of teaching. The mathematical model consists of a set of abstract axioms. The fundamental elements or terms of the axioms are undefined. These primitive or undefined terms obtain their definitions only implicitly through the axioms.

Manufacturers around the world are increasingly challenged to make components that are becoming smaller, more precise, more complex, and comprised of many more features. When manufactured components require precision, when they are very complex, or when they have multiple features, especially three dimensional features, manufacturers must often resort to machining. Machining smaller parts, however, can be particularly challenging. In this thesis, the issues associated with machining small parts are examined and a brief overview of the equipment available to machine small parts is considered. A machine tool design is developed which addresses many of the limitations associated with the Swiss-type screw machine. The proposed design is then "virtually prototyped" as a solid model in SolidWorksTM, machining forces are calculated, and the effects of the machining forces on the machine tool design are analyzed utilizing COSMOSWORKSTM FEA software and standard industrial formulas for calculating machine tool component service lives and safety factors. Values for all of the design metrics as well as all of the component service lives were found to meet or exceed their target values, thus the machine tool, if manufactured, is expected to perform robustly and to function as desired.

When there are some influential observations present in a data set (such as outliers or leverage points), the use of some robust method may be desirable for being able to draw relevant conclusions from an econometric analysis. In order to use these methods properly, we need some diagnostic tools. To be able to derive these tools theoretically, we first need to know the form of the asymptotic representation of corresponding estimator. This thesis derives the asymptotic representation of the estimator obtained by the method of least weighted squares under the assumption of heteroskedastic residuals. The tight- ness of the estimator and its asymptotic representation under several levels of contamination is also shown in a simulation study.

La Spectroscopie de Masse d'Ions Secondaires (SIMS) permet d'obtenir des images de distributions d'atomes à la surface d'un échantillon. La réponse impulsionnelle (RI) de l'instrument est inconnue. La déconvolution en aveugle a pour but d'enlever le flou associé. Ce problème mal conditionné est résolu en contraignant sa solution (régularisation). Le degré optimum de régularisation dépend d'un paramètre à déterminer. Il est trouvé, ainsi que ceux de la RI, par la méthode de validation croisée généralisée. Une étape de calibrage restreint l'espace de recherche des paramètres de la RI et les calculs sont accélérés en exploitant le modèle gaussien. L'image est déconvoluée en résolvant un grand système linéaire par la méthode du gradient conjugué. Un préconditionnement exploitant la séparabilité de la RI (isotrope ou anisotrope) en accélère la convergence. On montre comment utiliser plusieurs images d'un échantillon pour avoir une résolution plus fine (super-résolution). / Secondary Ion Mass Spectrometry (SIMS) creates images of atomic distributions on a sample's surface. The point spread function (PSF) is unknown. Blind deconvolution is used to remove the associated blur. This ill-conditionned problem is solved by constraining its solution (regularization). The optimum degree of regularization depends on a parameter to be determined. This parameter is found, as well as those of the PSF, by the generalized cross validation method. A calibration phase reduces the search space for the PSF parameters. The gaussian model used for the PSF is exploited to accelerate the computations. The image is deconvolved by solving a large linear system with the conjugate gradient method. A preconditionner making use of the PSF separability (isotropic or anisotropic) speeds up convergence.

Title: Option Pricing Author: Radek Moravec Department: Department of Probability and Mathematical Statistics Supervisor: doc. RNDr. Jan Hurt, CSc., Department of Probability and Mathematical Statistics In the present thesis we deal with European call option pricing using lattice approaches. We introduce a discrete market model and show a way how to find an arbitrage price of financial instruments on complete markets. It's equal to the discounted value of future expected cash flow. We present the binomial option pricing model and generalize it into multinomial model. We test the resulting formula on real market data obtained from NYSE and NASDAQ. We suggest a parameter estimate method which is based on time series of historical observations of daily close price. We compare calculated option prices with their real market value and try to explain the reasons of the differences. 1

In practice, we often meet data in the form of observations of several variables at various points in time. These data are called time series. We present various approaches in time series analysis; graphical models, vector autoregres- sive models and vector moving-average models. We try to get information about mutual relationship of the variables and then to model their behavior. The used techniques are illustrated on log returns of monthly average exchange rates. The programs are processed in the software Mathematica 7 and can be found on the CD. 1

The thesis describes a method for predicting avalanche activity in Scotland, based primarily on meteorological and topographical data, and this is incorporated into a framework for avalanche forecasting. A literature review identifies the main meteorological and topographical factors causing avalanche release. Snow accumulation, cold temperatures or thaws are identified as the major meteorological factors causing avalanches, and the thesis so distinguishes two basic avalanche types---the direct-action avalanche caused by fresh snow accumulation, and the climax avalanche caused by temperature changes. Slope angle, surface roughness, slope geometry and catchment area are identified as the major topographical controls on the location of avalanche activity. The thesis presents data on avalanche activity in Scotland over the last two hundred and eighty years, with particular emphasis on the Cairngorm Mountains. Avalanche locations, types and magnitudes for almost a thousand avalanches are described. Data on Scottish weather and terrain conditions are also presented, also with particular emphasis on the Cairngorm area. Major periods of snow accumulation, cold temperatures and thaws are identified between the winters of 1977/78 and 1979/80, and the location of steep slopes, smooth surfaces and large catchment areas in the Cairngorms is described. Data on the characteristics of the Scottish snow cover, based on work performed in the Cairngorms by E. Langmuir, B. Beattie and the author are then presented. Correlations between avalanche activity and meteorological, topographical and glaciological conditions are demonstrated, and this enables a predictive model to be developed which assesses avalanche probability on the basis of the amount of fresh snowfall and the prevailing temperature. Fresh snowfall of 200 mm., and either several days of cold weather with maximum temperatures below -4&deg;C or two or three days of warm weather with maximum temperatures above 0&deg;C lead to avalanche activity. The model can be continually up-dated and permits a continuous avalanche probability assessment to be made.

There are many multiple-model (MM) target-tracking algorithms that are available but there has yet to be a comparison that includes all of them. This work compares seven of the currently most popular MM algorithms in terms of performance, credibility, and computational complexity. The algorithms to be considered are the autonomous multiple-model algorithm, generalized pseudo- Bayesian of first order, generalized pseudo-Bayesian of second order, interacting multiple-model algorithm, B-Best algorithm, Viterbi algorithm, and reweighted interacting multiple-model algorithm. The algorithms were compared using three scenarios consisting of maneuvers that were both in and out of the model set. Based on this comparison, there is no clear-cut best algorithm but the B-best algorithm performs best in terms of tracking errors and the IMM algorithm has the best computational complexity among the algorithms that have acceptable tracking errors.