Introducción: Las recomendaciones del ATPIII y las SSCC europeas en prevención primaria cardiovascular aconsejan para la estimación del riesgo individual las funciones de Framingham y SCORE respectivamente. Ambas disponen de versiones calibradas para la población española. Todas, excepto Framingham(ATPIII), han sido derivadas para aplicarlas en un rango de edad limitado. Por otra parte, ninguna de estas funciones predice el riesgo cardiovascular total, ya que SCORE y SCORE calibrada sólo estima el riesgo de muerte y Framingham(ATPIII) y REGICOR el riesgo coronario. Sin embargo todas las medidas preventivas van dirigidas a la prevención de la enfermedad cardiovascular globalmente y las funciones son utilizadas como proxi del riesgo cardiovascular total. El objetivo es determinar cuál de estas funciones de estimación del riesgo cardiovascular es más útil para seleccionar al mayor número de pacientes con incidencia total de episodios cardiovasculares elevada. Método: Estudio observacional de cohortes de 1642 pacientes de 20 a 79 años no diabéticos, sin enfermedad cardiovascular establecida, en una Unidad de Hipertensión y Riesgo Vascular, seguidos un tiempo mediano de 3,8 años (rango 1- 10 años). Se registraron todos los episodios cardiovasculares mayores incidentes. Se calculó el riesgo predicho individual utilizando cada función. Se analizó el grado de concordancia para clasificar en grupos de riesgo utilizando los puntos de corte ≥5% en SCORE y SCORE calibrada, >20% para Framingham(ATPIII) y ≥10% en REGICOR. Se estimó para cada función la sensibilidad, especificidad, CPP, CPN, Odds ratio diagnóstica y exactitud. Se estimó la densidad de incidencia por el método de Mantel-Haenzel, y la incidencia acumulada en grupos de riesgo obtenidos al aplicar los algoritmos y compararlos mediante análisis de Kaplan-Meier. Se compararon las curvas de supervivencia mediante test de Breslow. Resultados: SCORE calibrada clasificó de alto riesgo a un 13,5%, SCORE a un 9,8%, Framingham(ATPIII) al 8,8% y REGICOR a un 3,5%. Durante el seguimiento 74 pacientes presentaron un primer episodio cardiovascular. La concordancia a la hora de clasificar en grupos de riesgo entre Framingham y SCORE (Kappa=0,37), y entre SCORE calibrada y REGICOR (Kappa=0,34) fue débil. Para cada función, el grupo de alto riesgo presentó un HR superior a 4 respecto al grupo de riesgo moderado/bajo. Todas las funciones presentaron una baja sensibilidad sin diferencias significativas, así como una exactitud moderada para discriminar pacientes con episodio cardiovascular. Al comparar SCORE vs Framingham(ATPIII) y SCORE calibrada vs REGICOR, la incidencia de episodios cardiovasculares en los grupos con discordancia en la clasificación de alto riesgo no presentó diferencias significativas. SCORE excluyó un 4,8% de pacientes de alto riesgo frente a un 5,8% por Framingham(ATPIII). SCORE calibrada excluyó un 0,4% de pacientes de alto riesgo frente a un 10,3% en el caso de REGICOR. Además, SCORE calibrada seleccionó un 4,2% más de pacientes respecto a SCORE con una incidencia significativamente más elevada que el grupo de riesgo moderado/bajo. En las mujeres por falta de potencia estadística no se observaron diferencias significativas. En los hombres SCORE excluyó un 7,1% de pacientes de alto riesgo frente a un 10,6% por Framingham(ATPIII), mientras que SCORE calibrada excluyó un 0,4% de pacientes de alto riesgo frente a un 16,1% por REGICOR con una probabilidad acumulada de ECV a los 5 años del 19%. Además, SCORE calibrada incluyó un 5% más de pacientes que SCORE con una probabilidad acumulada de episodios cardiovasculares a los 5 años del 17%. Conclusión: En prevención primaria, aunque el rendimiento de las cuatro funciones analizadas no presenta diferencias significativas, la concordancia entre ellas es débil, siendo SCORE calibrada la más útil para seleccionar al mayor número de pacientes con una incidencia de episodios cardiovasculares aumentada. / Background: The recommendations of ATPIII and the European scientific societies in cardiovascular primary prevention advise the estimate of the individual risk by Framingham and SCORE functions respectively. Both have versions calibrated for the Spanish population. The objective is to determine which of these functions is more useful to select to the greater number of patients with high cardiovascular incidence. Method: Longitudinal study of 1642 non-diabetic patients, without previous history of cardiovascular disease attending to a Hypertension and Vascular clinic. High-risk patients were considered those with ≥5% risk according to SCORE and calibrated SCORE, >20% on Framingham(ATPIII) and ≥10% on REGICOR. Major cardiovascular events were collected during a 3 year median follow-up (range 1-10 años) and analysed by Kaplan-Meier. Results: 13,5% of patients was considered of high-risk with calibrated SCORE function, 9,8& with SCORE, 8,8% with Framingham(ATPIII) and 3,5% with REGICOR. The concordance between SCORE and Framingahm(ATPIII) (Kappa=0,37), and between calibrated SCORE and REGICOR (Kappa=0,34) was poor. During follow-up 74 cardiovascular events occurred. The incidence of cardiovascular events in groups with disagreement in high risk classification, did not present differences, upon compared SCORE vs Framingham(ATPIII) and calibrated SCORE vs REGICOR. SCORE excluded a 4,8% of high risk patients as compared to a 5,8% by Framingham(ATPIII). Calibrated SCORE excluded a 0,4% of high risk patients as compared to a 10,3% in the case of REGICOR. Furthermore, calibrated SCORE selected 4,2% patients more than SCORE with incidence of cardiovascular events significantly highest that moderate/low risk group. In women, by lack of statistics power, it’s not observed meaningful differences. In the men SCORE excluded 7,1% of high risk patients as compared to 10,6% by Framingham(ATPIII). Calibrated SCORE excluded 0,4% of high risk patients as compared to 16,1% by REGICOR with acumulative probability of cardiovascular events at 5 years of 19%. Furthermore, calibrated SCORE included a 5% of men more than SCORE with an acumulative probability of cardiovascular events at 5 years of 17%. Conclusions: In primary prevention, calibrated SCORE is the most useful to select the greater number of patients with a cardiovascular events incidence increased.

Accurately and efficiently characterizing the material properties of hot mix asphalt (HMA) is critical to the design and development of pavements that can experience repeated loading for long periods of time and resist fatigue cracking. The Calibrated Mechanistic with Surface Energy (CMSE) method of design to preclude this primary type of distress requires that the HMA material be tested using the Relaxation Modulus (RM) and Repeated Direct Tension (RDT) tests to determine the material properties required for accurate calculations. The RM test requires considerable time to complete and provides results with relatively high variability. Further research has lead to the development of the Viscoelastic Characterization (VEC) test, from which the RM master curve can be developed. Material properties from the RM master curve can be easily determined and applied in the CMSE method. The modified repeated direct tension (RDT*) test removes rest periods and unwanted healing from the RDT test. The RDT* test also allows the dissipated pseudo strain energy (DPSE) to be separated into permanent deformation and fatigue cracking energies. The rate of change in DPSE associated with fatigue can then be applied in the CMSE method. Data sets for these tests are extensive and time consuming to analyze. Microsoft Excel spreadsheet macros were developed to reduce the time required for analysis from an estimated 10 hours to approximately 8 minutes. Testing of 14 different samples showed that the VEC and RDT* tests still required some adjustments in order to get accurate results. The rate of loading in the VEC test must be reduced to allow sufficient testing time to obtain the required data. The RDT* test requires a decrease in the controlling strain levels from 80 mu-epsilon and 350 mu-epsilon to 20 mu-epsilon and 175 mu-epsilon for the undamaged and damaged portions of the test, respectively. Testing of a sample using the new VEC and RDT* test recommendations showed that the recommended changes provided better results. Samples were undamaged where required and damaged portions of the test ran to completion without causing compression or sample failure. Material properties can be accurately determined and applied in the CMSE method.

E-learning tools such as Calibrated Peer Review (CPR) have made writing assignments easier to implement and grade; however, we have limited knowledge of how CPR affects student scientific writing. Past CPR research has examined how CPR generated scores change across multiple CPR writing assignments for the purpose of reporting student learning gains. This study will not rely on CPR generated score data. This study (1) independently evaluated the impact of CPR on student writing of ecological inquiry report components using a grading criteria instrument and (2) explored how the revision process influenced the quality of ecological inquiry report components through text analysis. A web-based science inquiry project was implemented in a large (up to 500 students) introductory ecology course. Students observed grizzly bears at McNeil River Falls in Alaska using Bear Cam picture stills. They developed and tested hypotheses about grizzly bear spatial distribution and interactions and reported findings in individual ecological inquiry reports. Students submitted reports to CPR and anonymously reviewed three peer reports and self-assessed their own. Finally, students were given one-week following CPR to revise reports based on peer reviews and submit online. A 28-item grading criteria instrument (9 scales) was used to examine how students revised ecological inquiry reports post CPR. Eight paired t-tests were used to assess the pre-post CPR changes in scores for individual grading criteria scales or components. Cohen's d effect size was used to explore how achievement or performance level, ethnicity, gender and major influenced student text changes to ecological inquiry report components post CPR. Text analysis using a subset of 27 sample reports (pre-post CPR) assessed the amount and location of text changes and the impact of these revisions on the quality of ecological inquiry report components. Common errors in ecological inquiry report components post CPR were also analyzed. Results showed that CPR and revision significantly improved the scores related to the objective, sampling and discussion scales. Analyses using Cohen's d effect sizes illustrated interesting but inconsistent patterns related to the influence of student performance level, gender, ethnicity, and major on pre-post CPR score gains. Text analysis revealed the majority of helpful revisions were related to making the objective identifiable, reporting of sample size and discussion of study limitations and future questions raised by individual ecological inquiry projects. Text analysis shows three common reasons participants failed to meet grading criteria post CPR. Un-testable hypotheses, insufficient descriptions for sample selection, data analysis, variables collected and revisions of only easy grading criteria components. This study provided direct evidence of CPR's effects on student writing and provided a greater understanding of pattern of revision process following CPR.

No description available.

When a company evaluates a customer for being a potential prospect, one of the key questions to answer is whether the customer will generate profit in the long run. A possible step to answer this question is to predict the likelihood of the customer returning to the company again after the initial purchase. The aim of this master thesis is to investigate the possibility of using machine learning techniques to predict the likelihood of a new customer returning for a second purchase within a certain time frame. To investigate to what degree machine learning techniques can be used to predict probability of return, a number of di↵erent model setups of Logistic Lasso, Support Vector Machine and Extreme Gradient Boosting are tested. Model development is performed to ensure well-calibrated probability predictions and to possibly overcome the diculty followed from an imbalanced ratio of returning and non-returning customers. Throughout the thesis work, a number of actions are taken in order to account for data protection. One such action is to add noise to the response feature, ensuring that the true fraction of returning and non-returning customers cannot be derived. To further guarantee data protection, axes values of evaluation plots are removed and evaluation metrics are scaled. Nevertheless, it is perfectly possible to select the superior model out of all investigated models. The results obtained show that the best performing model is a Platt calibrated Extreme Gradient Boosting model, which has much higher performance than the other models with regards to considered evaluation metrics, while also providing predicted probabilities of high quality. Further, the results indicate that the setups investigated to account for imbalanced data do not improve model performance. The main con- clusion is that it is possible to obtain probability predictions of high quality for new customers returning to a company for a second purchase within a certain time frame, using machine learning techniques. This provides a powerful tool for a company when evaluating potential prospects.

Vehicle automation has garnered increasing attention as a means of improving safety and efficiency. Advanced Driver Assistance Systems (ADAS) have gained popularity in the transport industry. However, establishing an appropriate level of trust in these systems is crucial for their successful implementation. This research explores the factors influencing driver trust calibration in different levels of automation within driver assistance systems for commercial mobility trucks to ensure drivers comprehend the limitations of these systems and uphold road safety. A qualitative approach involved eleven interviews and observations with drivers to explore their perceptions, experiences, and expectations regarding these systems. The study’s findings extend the Hoff and Bashir Trust model to include significant social factors in calibrating trust. These findings offer valuable insights into the various trust factors that impact driver trust calibration at different levels of automation in driver assistance systems for commercial mobility trucks. These insights contribute to academia in that they help understand how trust in automation is formed and calibrated in real-world settings. In the automotive industry, they can guide the design and implementation of these systems to enhance future drivers’ safety and overall experience.

This dissertation proposes a set of methods for improving spatial localization of cerebral metabolic changes using functional magnetic resonance imaging (fMRI). Blood oxygen level dependent (BOLD) fMRI estabilished itself as the most frequently used technique for mapping brain activity in humans. It is non-invasive and allows to obtain information about brain oxygenation changes in a few minutes. It was discovered in 1990 and, since then, it contributed enormously to the developments in neuroscientific research. Nevertheless, the BOLD contrast suffers from inherent limitations. This comes from the fact that the observed response is the result of a complex interplay between cerebral blood flow (CBF), cerebral blood volume (CBV) and cerebral metabolic rate of oxygen consumption (CMRO2) and has a strong dependency on baseline blood volume and oxygenation. Therefore, the observed response is mislocalized from the site where the metabolic activity takes place and it is subject to high variability across experiments due to normal brain physiology. Since the peak of BOLD changes can be as much as 4 mm apart from the site of metabolic changes, the problem of spatial mislocalization is particularly constraining at submillimeter resolution. Three methods are proposed in this work in order to overcome this limitation and make data more comparable. The first method involves a modification of an estabilished model for calibration of BOLD responses (the dilution model), in order to render it applicable at higher resolutions. The second method proposes a model-free scaling of the BOLD response, based on spatial normalization by a purely vascular response pattern. The third method takes into account the hypothesis that the cortical vasculature could act as a low-pass filter for BOLD fluctuations as the blood is carried downstream, and investigates differences in frequency composition of cortical laminae. All methods are described and tested on a depth-dependent scale in the human motor cortex.

No description available.

This study examined the feasibility of using Rasch true score preequating under violated model assumptions and nonequivalent populations. Dichotomous item responses were simulated using a compensatory two dimensional (2D) three parameter logistic (3PL) Item Response Theory (IRT) model. The Rasch model was used to calibrate difficulty parameters using two methods: Fixed Parameter Calibration (FPC) and separate calibration with the Stocking and Lord linking (SCSL) method. A criterion equating function was defined by equating true scores calculated with the generated 2D 3PL IRT item and ability parameters, using random groups equipercentile equating. True score preequating to FPC and SCSL calibrated item banks was compared to identity and Levine's linear true score equating, in terms of equating bias and bootstrap standard errors of equating (SEE) (Kolen & Brennan, 2004). Results showed preequating was robust to simulated 2D 3PL data and to nonequivalent item discriminations, however, true score equating was not robust to guessing and to the interaction of guessing and nonequivalent item discriminations. Equating bias due to guessing was most marked at the low end of the score scale. Equating an easier new form to a more difficult base form produced negative bias. Nonequivalent item discriminations interacted with guessing to magnify the bias and to extend the range of the bias toward the middle of the score distribution. Very easy forms relative to the ability of the examinees also produced substantial error at the low end of the score scale. Accumulating item parameter error in the item bank increased the SEE across five forms. Rasch true score preequating produced less equating error than Levine's true score linear equating in all simulated conditions. FPC with Bigsteps performed as well as separate calibration with the Stocking and Lord linking method. These results support earlier findings, suggesting that Rasch true score preequating can be used in the presence of guessing if accuracy is required near the mean of the score distribution, but not if accuracy is required with very low or high scores.

Šiame magistro darbe nagrinėjamos papildomos informacijos panaudojimo galimybės konstruojant baigtinės populiacijos koreliacijos koeficiento įvertinius. Taikant imties plano svorių kalibravimo metodus sukonstruojami aštuoni nauji koreliacijos koeficiento įvertiniai. Įvertiniams konstruoti panaudotos kelios kalibravimo lygčių ir atstumo funkcijų kombinacijos. Kalibruotieji svoriai išvesti pasitelkiant Lagranžo neapibrėžtųjų daugiklių metodą. Naudojant Teiloro ištiesinimo ir visrakčio metodus sukonstruojami koreliacijos koeficiento įvertinių dispersijos įvertiniai. Darbe taip pat atliekamas matematinis modeliavimas, kuriuo siekiama palyginti naujai sukonstruotus įvertinius tarpusavyje ir su standartiniu įvertiniu. Įvertinių tikslumą lemia pakankamai didelė koreliacija tarp tyrimo kintamųjų ir jų papildomų kintamųjų, o du įvertiniai yra tikslūs ir esant mažai koreliacijai. Tiriama, kaip įvertinių tikslumą įtakoja imties dydis bei koreliacija tarp tyrimo ir papildomų kintamųjų. Matematinio modeliavimo eksperimentai atlikti, naudojant darbo autorės sukurtas MATLAB programas. / This master’s thesis analyzes the opportunities of auxiliary information usage while constructing the correlation coefficient estimators of the finite population. By applying the methods of weight calibration of sample design the new eight correlation coefficient estimators are derived. Several combinations of calibration equations and distance measures are used to construct estimators. Calibration weights were derived through Lagrange multiplier method. Using Taylor linearization and jackknife methods variance estimators of the correlation coefficient estimators were constructed. The paper also contains a mathematical simulation aimed at comparing the new derived estimators in between and with a standard estimator. The accuracy of estimators is determined by rather substantial correlation between study and auxiliary variables, whereas two of the estimators are accurate even in the presence of bad correlation. It is tested how sample size and correlation between study and auxiliary variables influences the accuracy of estimators. All the mathemathical simulation tests were performed employing MATLAB programs that have been created by the author of this work.