Assessment of the effect on statistical power of regression model misspecification by using techniques of mathematical statistics and simulation study

Objectives. This paper seeks to assess the effect on statistical power of regression model misspecification in a variety of situations. ^ Methods and results. The effect of misspecification in regression can be approximated by evaluating the correlation between the correct specification and the misspecification of the outcome variable (Harris 2010).In this paper, three misspecified models (linear, categorical and fractional polynomial) were considered. In the first section, the mathematical method of calculating the correlation between correct and misspecified models with simple mathematical forms was derived and demonstrated. In the second section, data from the National Health and Nutrition Examination Survey (NHANES 2007-2008) were used to examine such correlations. Our study shows that comparing to linear or categorical models, the fractional polynomial models, with the higher correlations, provided a better approximation of the true relationship, which was illustrated by LOESS regression. In the third section, we present the results of simulation studies that demonstrate overall misspecification in regression can produce marked decreases in power with small sample sizes. However, the categorical model had greatest power, ranging from 0.877 to 0.936 depending on sample size and outcome variable used. The power of fractional polynomial model was close to that of linear model, which ranged from 0.69 to 0.83, and appeared to be affected by the increased degrees of freedom of this model.^ Conclusion. Correlations between alternative model specifications can be used to provide a good approximation of the effect on statistical power of misspecification when the sample size is large. When model specifications have known simple mathematical forms, such correlations can be calculated mathematically. Actual public health data from NHANES 2007-2008 were used as examples to demonstrate the situations with unknown or complex correct model specification. Simulation of power for misspecified models confirmed the results based on correlation methods but also illustrated the effect of model degrees of freedom on power.^

FACTORS INFLUENCING THE PATTERNS OF PHYSICIAN VISITS TO NURSING HOME RESIDENTS

This study was designed to identify some of the factors related to patterns of physician visits to nursing home residents. The relationship of ten resident and organizational characteristics to patterns of physician visits was investigated through secondary analysis of data abstracted from the 1973-74 National Nursing Home Survey of the National Center for Health Statistics. The study sample was composed of 11,135 of the 19,013 nursing home residents who participated in the survey.^ The analytic results revealed that all ten variables had a statistically significant relationship to patterns of physician visits, mainly due to the large sample size. The degrees of association between the variables, measured by the Cramer's V statistic, ranged from moderate to very weak.^ Certification status of the nursing home under Medicare and/or Medicaid was shown to be most strongly related to patterns of physician visits, followed by primary source of payment for nursing home care, and residence prior to nursing home admission. Several variables thought to be related to patterns of physician visits were found to have a very weak relationship: age of the resident, marital status, length of stay, primary diagnosis, number of chronic conditions, activities of daily living status, and levels of care.^ In order to get a more precise picture of the relative influence of certification status and primary source of payment when the other variables were statistically controlled, these two variables were combined into a single variable. The results revealed that the combined effects of certification status and primary source of payment were sustained, regardless of differences in the residents' personal, utilization, and health status characteristics, and the levels of care that they received. The results also indicated that the five groups created by combining the two variables differed in patterns of physician visits. For example, private pay residents in intermediate care facilities (ICF's) and non-certified facilities were more likely to receive unscheduled visits than private pay residents in skilled nursing homes (SNH's), residents in SNH's supported by Medicare or Medicaid, and residents in ICF's supported by Medicaid. ^

Detecting genetic and nutritional lung cancer risk factors related to folate metabolism using Bayesian generalized linear models

Complex diseases, such as cancer, are caused by various genetic and environmental factors, and their interactions. Joint analysis of these factors and their interactions would increase the power to detect risk factors but is statistically. Bayesian generalized linear models using student-t prior distributions on coefficients, is a novel method to simultaneously analyze genetic factors, environmental factors, and interactions. I performed simulation studies using three different disease models and demonstrated that the variable selection performance of Bayesian generalized linear models is comparable to that of Bayesian stochastic search variable selection, an improved method for variable selection when compared to standard methods. I further evaluated the variable selection performance of Bayesian generalized linear models using different numbers of candidate covariates and different sample sizes, and provided a guideline for required sample size to achieve a high power of variable selection using Bayesian generalize linear models, considering different scales of number of candidate covariates. ^ Polymorphisms in folate metabolism genes and nutritional factors have been previously associated with lung cancer risk. In this study, I simultaneously analyzed 115 tag SNPs in folate metabolism genes, 14 nutritional factors, and all possible genetic-nutritional interactions from 1239 lung cancer cases and 1692 controls using Bayesian generalized linear models stratified by never, former, and current smoking status. SNPs in MTRR were significantly associated with lung cancer risk across never, former, and current smokers. In never smokers, three SNPs in TYMS and three gene-nutrient interactions, including an interaction between SHMT1 and vitamin B12, an interaction between MTRR and total fat intake, and an interaction between MTR and alcohol use, were also identified as associated with lung cancer risk. These lung cancer risk factors are worthy of further investigation.^

Weight-for-age percentile and its association to community acquired methicillin resistant Staphylococcus aureus among pediatric patients in Houston, TX

BACKGROUND: Weight has been implicated as a risk factor for symptomatic community-acquired methicillin resistant Staphylococcus Aureus (CA-MRSA). Information from Texas Children's Hospital (TCH) in Houston, TX was used to implement a case-control study to assess weight-for-age percentile (WFA), race and seasonal exposure as risk factors. ^ METHODS: A retrospective chart review to collect data from TCH was conducted covering the time period January 1st, 2008 to May 31st, 2011. Cases were confirmed and identified by the infectious disease department and were matched on a 1:1 ratio to controls that were seen by the emergency department for non-infected fractures from June 1st, 2008 to May 31st, 2011. Data abstraction was performed using TCH's electronic medical records (EMR) system (EPIC ®). ^ RESULTS: Of 702 CA-MRSA identified cases, ages 9 to 16.99, 564 (80.3%) had the variable `weight' present in their EMR, were not duplicates and not determined to be outliers. Cases were randomly matched to a pool of available controls (n=1864) according to age and gender, yielding 539 1:1 matched pairs (95.5% case matching success) with a total study sample size, N=1078. Case median age was 13.38 years with the majority being White (66.05%) and male (59.4%). Adjusted conditional logistic regression analysis of the matched pairs identified the following risk factors to presenting with CA-MRSA infection among pediatric patients, ages 9 to 16.99 years: a) Individual weight in the highest (75th-99.9th) WFA quartile (OR=1.36; 95% confidence interval [CI]=1.06-1.74; P= 0.016), b) Infection during summer months (OR: 1.69; 95% CI=1.2-2.38; P= 0.003), c) patients of African American race/ethnicity (OR= 1.48; 95% CI=1.13-1.95; P= 0.004). ^ CONCLUSIONS: Pediatric patients, 9 to 16.99 years of age, in the highest WFA quartile (75th-99.9th), or of African-American race had an associated increased risk of presenting with CA-MRSA infection. Furthermore, children in this population were at a higher risk of contracting CA-MRSA infection during the summer season.^

Decomposition of $\rm R\times C$ contingency table chi -square: Application to binned DNA fragment size data and population structure analysis

The purpose of this research is to develop a new statistical method to determine the minimum set of rows (R) in a R x C contingency table of discrete data that explains the dependence of observations. The statistical power of the method will be empirically determined by computer simulation to judge its efficiency over the presently existing methods. The method will be applied to data on DNA fragment length variation at six VNTR loci in over 72 populations from five major racial groups of human (total sample size is over 15,000 individuals; each sample having at least 50 individuals). DNA fragment lengths grouped in bins will form the basis of studying inter-population DNA variation within the racial groups are significant, will provide a rigorous re-binning procedure for forensic computation of DNA profile frequencies that takes into account intra-racial DNA variation among populations. ^

Nd, Sr and Pb isotopic composition of eolian dust deposited in the central North Pacific

The silicate fractions of recent pelagic sediments in the central north Pacific Ocean are dominated by eolian dust derived from central Asia. An 11 Myr sedimentary record at ODP Sites 885/886 at 44.7°N, 168.3°W allows the evaluation of how such dust and its sources have changed in response to late Cenozoic climate and tectonics. The extracted eolian fraction contains variable amounts (>70%) of clay minerals with subordinate quartz and plagioclase. Uniform Nd isotopic compositions (epsilon-Nd =38.6 to 310.5) and Sm/Nd ratios (0.170-0.192) for most of the 11 Myr record demonstrate a well-mixed provenance in the basins north of the Tibetan Plateau and the Gobi Desert that was a source of dust long before the oldest preserved Asian loess formed. epsilon-Nd values of up to 36.5 for samples 62.9 Ma indicate <=35 wt% admixture of a young, Kamchatka-like volcanic arc component. The coherence of Pb and Nd in the erosional cycle allows us to constrain the Pb isotopic composition of Asian loess devoid of anthropogenic contamination to 206Pb/204Pb =18.97 +/- 0.06, 207Pb/204Pb =15.67 +/- 0.02, 208Pb/204Pb =39.19 +/- 0.11. 87Sr/86Sr (0.711-0.721) and Rb/Sr ratios (0.39-1.1) vary with dust mineralogy and provide an age indication of ~250 Ma. 40Ar/39Ar ages of six dust samples are uniform around 200 Ma and match the K-Ar ages of modern dust deposited on Hawaii. These data reflect the weighted age average of illite formation. Changes from illite- smectite with significant kaolinite to illite- and chlorite-rich, kaolinite-free assemblages since the late Pliocene document changes in the intensity of chemical weathering in the source region. Such weathering evidently did not disturb the K-Ar systematics, and only induced scatter in the Rb-Sr data. We propose that when smectite forms at the expense of illite, K and Ar are quantitatively lost from what becomes smectite, but are quantitatively retained in adjacent illite layers. 40Ar/39Ar age data, therefore, are insensitive to smectite formation during chemical weathering but date the diagenetic growth of illite, the major K-bearing phase in the dust. Over the past 12 Myr, the dust flux to the north Pacific increased by more than an order of magnitude, documenting a substantial drying of central Asia. This climatic change, however, did not alter the ultimate source of the dust, and neoformational products of chemical weathering always remained subordinate to assemblages reworked by mechanical erosion in dust deposited in eastern Asia and the Pacific Ocean.

Chemical and germanium isotopic compositions of sediments and sedimentary rocks from ODP Sites 129-801 and 185-1149

A new technique for the precise and accurate determination of Ge stable isotope compositions has been developed and applied to silicate rocks and biogenic opal. The analyses were performed using a continuous flow hydride generation system coupled to a MC-ICP-MS. Samples have been purified through anion- and cation-exchange resins to separate Ge from matrix elements and eliminate potential isobaric interferences. Variations of 74Ge/70Ge ratios are expressed as d74Ge values relative to our internal standard and the long-term external reproducibility of the data is better than 0.2? for sample size as low as 15 ng of Ge. Data are presented for igneous and sedimentary rocks, and the overall variation is 2.4? in d74Ge, representing 12 times the uncertainty of the measurements and demonstrating that the terrestrial isotopic composition of Ge is not unique. Co-variations of 74Ge/70Ge, 73Ge/70Ge and 72Ge/70Ge ratios follow a mass-dependent behaviour and imply natural isotopic fractionation of Ge by physicochemical processes. The range of d74Ge in igneous rocks is only 0.25? without systematic differences among continental crust, oceanic crust or mantle material. On this basis, a Bulk Silicate Earth reservoir with a d74Ge of 1.3+/-0.2? can be defined. In contrast, modern biogenic opal such as marine sponges and authigenic glauconite displayed higher d74Ge values between 2.0? and 3.0?. This suggests that biogenic opal may be significantly enriched in light isotopes with respect to seawater and places a lower bound on the d74Ge of the seawater to +3.0?.This suggests that seawater is isotopically heavy relative to Bulk Silicate Earth and that biogenic opal may be significantly fractionated with respect to seawater. Deep-sea sediments are within the range of the Bulk Silicate Earth while Mesozoic deep-sea cherts (opal and quartz) have d74Ge values ranging from 0.7? to 2.0?. The variable values of the cherts cannot be explained by binary mixing between a biogenic component and a detrital component and are suggestive of enrichment in the light isotope of diagenetic quartz. Further work is now required to determine Ge isotope fractionation by siliceous organisms and to investigate the effect of diagenetic processes during chert lithification.

Recursos intangibles como factores clave en la gestión estratégica de instituciones de educación superior y su impacto en los procesos de innovación

La presente investigación, tiene como objetivo analizar las influencias que ejercen los recursos intangibles (Gestión del Conocimiento, Marca, Reputación Organizacional y Responsabilidad Social) en la gestión estratégica de las instituciones de educación superior (IES) y el impacto de los mismos en los procesos de innovación a través del valor añadido que se transfiere al entorno. Se considera importante realizar un estudio sobre este tema dado que son las IES las encargadas de proporcionar los conocimientos y los nuevos hallazgos en innovaciones tecnológicas, que son transferidas al tejido productivo de las regiones, lo que proporciona crecimiento económico y mejoras en la calidad de vida. El estudio se enmarca dentro de los postulados de la teoría de los recursos y las capacidades (TRC) y de los intangibles, los cuales sirven de base a la investigación. Se planteó un sistema de hipótesis subdividido en dos vías de influencias. La primera, donde se analizan las influencias directas que ejercen los recursos intangibles sobre los resultados de las IES. La otra vía es la indirecta, que estudia las influencias que ejercen los recursos intangibles gestionados estratégicamente sobre los resultados de las IES. Esta investigación se ha concebido como no experimental, de tipo exploratorio, basada en el paradigma que busca explicar un fenómeno (variable dependiente) a través del comportamiento de las variables independientes. Es un estudio transversal, cuantitativo, que intenta describir las causas del fenómeno. Con el objeto de determinar las influencias o relaciones de causalidad que subyacen entre las variables, se utilizó la técnica del modelo de ecuaciones estructurales (SEM). La población objeto de estudio estuvo constituida por los 857 individuos pertenecientes a los consejos directivos de las IES, que forman parte de las base de datos que gestiona el Consorcio de Escuelas de Ingeniería de Latinoamérica y del Caribe y la Universidad Politécnica de Madrid, con un tamaño de muestra significativa de 250 directivos, lo que representa el 29,42% de la población. Como fuentes de recolección de información se utilizaron fuentes primarias y secundarias. Para recabar la información primaria se diseñó un cuestionario (ad hoc), el cual fue validado por expertos. La información de fuentes secundarias se extrajo de la bases de datos de la Red Iberoamericana de Ciencia y Tecnología (RICYT). Los resultados obtenidos indican que las influencias directas que pueden ejercer los recursos intangibles (Gestión del Conocimiento, Marca, Reputación Organizacional y Responsabilidad Social) no son significativas, por ello se rechazaron todas las hipótesis de la vía de influencia directa. Asimismo, de acuerdo con el contraste realizado al submodelo que representa la vía de influencia indirecta, resultaron significativas las influencias que ejercen los intangibles Gestión del Conocimiento y Reputación Organizacional, gestionadas estratégicamente sobre los resultados con valor añadido generado por las IES y transferidos al entorno. Sin embargo, no se apoyan todas las hipótesis del modelo, debido a que los constructos Marca y Responsabilidad Social resultaron no significativos. Las teorías sobre intangibles enmarcadas en la TRC no son del todo robustas y requieren de mayores esfuerzos por parte de los investigadores para lograr definir los constructos a utilizar. De igual forma, se sigue corroborando el desfase que existe entre las teorías que sustentan la investigación y las comprobaciones empíricas de las mismas. Además, se evidencia que las IES enfocan su actuación hacia la academia, por encima de las otras funciones, otorgando a la enseñanza e investigación y a la reputación organizacional una mayor importancia. Sin embargo, debido a su naturaleza no empresarial, las IES siguen manteniendo una filosofía de gestión enfocada a la generación y transmisión de conocimientos que crean reputación. Se excluyen los intangibles Marca y Responsabilidad Social, por considerar que no aportan valor a sus procesos internos o que están inmersos dentro de otros recursos intangibles. En conclusión, se corrobora el atraso de la gestión estratégica que presentan las IES en Latinoamérica. Se comprueba la no aplicación de postulados básicos de la gerencia moderna que contribuyan al manejo eficiente de todos sus recursos y al logro de sus objetivos. Esto deriva en la necesidad de modernizar la visión estratégica de las IES y en crear mejores mecanismos para lograr reconocer, mantener, proteger y desarrollar los Recursos Intangibles que poseen, realizando combinaciones de recursos óptimas, que maximicen la creación de valor para sí mismas y para la sociedad a la que pertenecen. ABSTRACT This research aims to analyze the influences exerted by intangible resources (Knowledge Management, Brand, Organizational Reputation and Social Responsibility) in the strategic management of higher education institutions (HEIs) and their impact in the innovation processes through the added value that is transferred to the environment. It is considered important to conduct a study on this issue since HEIs are responsible for providing knowledge and new findings on technological innovations, which are then, transferred to the productive fabric of these regions, providing economic growth and improvements in quality of life. The study is framed within the tenets of the Theory of Resources and Capabilities (TRC) and of intangibles which underlie this research. A system of hypotheses was raised which was subdivided into two pathways of influences. In the first system the direct influences exerted by intangible resources on the results of the IES are analyzed. The other system focuses on the indirect influences exerted by the strategically managed intangible resources on the HEIs results. This research is designed as experimental, exploratory and based on the paradigm that seeks to explain a phenomenon (the dependent variable) through the behavior of the independent variables. It is a crosssectional, quantitative study, which attempts to describe the causes of the phenomenon. In order to determine the influences or causal relationships among variables the structural equation modeling technique (SEM) was used. The population under study consisted of 857 individuals from the boards of HEIs, which are part of the database managed by the Consortium of Engineering Schools in Latin America and the Caribbean and the Technical University of Madrid, with a significant sample size of 250 managers which represents 29.42% of the population. As sources of information gathering primary and secondary sources were used. To collect primary information an ad-hoc questionnaire which was validated by experts was designed. The secondary information was extracted from the database of the Latin American Network of Science and Technology (RICYT). The results obtained indicate that the direct influences that intangible resources (Knowledge Management, Brand, Organizational Reputation and Social Responsibility) can exert are not significant. Therefore, all hypotheses related to direct influence were rejected. Also, according to the test made with the system which represents the indirect channel of influence, significant influences were exerted on the results with added value generated by the HEIs by the intangibles Knowledge Management and Organizational Reputation when they were managed strategically. However, all model hypotheses are not supported, because the constructs Brand and Social Responsibility were not significant. Theories of intangibles within the framework of the Theory of Resources and Capabilities are not entirely robust and require greater efforts by researchers to define the constructs to be used. Similarly the existing gap between the theories underpinning research and the empirical tests continues to be corroborated. In addition, there is evidence that HEIs focus their action on the academy neglecting the other functions, giving more importance to teaching, research and organizational reputation. However, due to their non-business nature, HEIs still maintain a management philosophy focused on the generation and transmission of knowledge which leads to reputation. The intangibles Brand and Social Responsibility are excluded, considering that they do not add value to their internal processes or are embedded within other intangible resources. In conclusion, the backwardness of HEIs’ strategic management in Latin America is confirmed. The lack of application of the basic principles of modern management that contribute to the efficient administration of all the resources and the achievement of objectives is proven. This leads to the need to modernize the strategic vision of HEIs and the need for better mechanisms to recognize, maintain, protect and develop the intangible resources they possess, achieving optimal combinations of resources in order to maximize the creation of value for them and for the society to which they belong.

Paracetamol vs ibuprofeno para el dolor de la episiotomía en las 42 horas postparto

Objetivo: Evaluación de la eficacia analgésica para el dolor de la episiotomía entre el paracetamol y el Ibuprofeno, en las primeras 42 horas postparto. Método: Estudio cuasi-experimental (prospectivo y simple ciego) en mujeres que dieron a luz en el HOSPITAL UNIVERSITARIO CENTRAL DE ASTURIAS (OVIEDO), excluyendo alérgicas, patologías asociadas ó aquellas que el idioma impidiese un correcto entendimiento. Dos grupos: 1) Paracetamol 1 gr; 2) Ibuprofeno 600 mg. Tamaño de muestra: 110 por grupo para alcanzar mínimo de 80. Variable principal: grado de dolor según puntuación de escala (0 a 3). Otras variables: edad de paciente, semanas de gestación, peso neonatal, paridad, inicio del parto, anestesia epidural, tipo de parto, desgarro, inflamación y enrojecimiento, hematoma, hemorroides, necesidad de sondaje evacuador, aplicación de hielo y solicitud de analgesia. Tamaño final de la muestra: 88 grupo paracetamol y 97 grupo ibuprofeno. La escala de dolor se midió a las 2 horas postparto (previo al tratamiento) y, posteriormente, cada 8 hasta 42 horas. Se realizó análisis descriptivo y comparación entre grupos. Resultados: No encontramos diferencias significativas en la escala de dolor entre ambos fármacos, ni en los subgrupos analizados, salvo en el subgrupo de partos eutócicos, donde el ibuprofeno fue superior al paracetamol. En el global de la serie, el grupo de paracetamol solicitó hielo y otra medicación con mayor frecuencia que el grupo de ibuprofeno. Conclusiones: El ibuprofeno 600 mg y el paracetamol de 1 gr obtienen una respuesta similar en las primeras 42 horas postparto, si bien el ibuprofeno parece tener algunas ventajas adicionales.

Paleomagnetic and rock magnetic properties of ODP Leg 173 sites, west Iberia continental margin

We present detailed paleomagnetic and rock magnetic results of rock samples recovered during Leg 173. The Leg 173 cores display a multicomponent magnetization nature. Variations in magnetic properties correlate with changes in lithology that result from differences in the abundance and size of magnetic minerals. The combined investigation suggests that the magnetic properties of the "fresher" peridotite samples from Site 1070 are controlled mainly by titanomagnetite, with a strong Verwey transition in the vicinity of 110 K, and with field- and frequency-dependent susceptibility curves that resemble those of titanomagnetites. These results are in excellent agreement with thermomagnetic characteristics where titanomagnetites with Curie temperature ~580°C were identified from the "fresher" peridotites. In contrast to the magnetic properties observed from the "fresher" peridotites, the low-temperature curves for the "altered" peridotites did not show any Verwey transition. Thermomagnetic analysis using the high-temperature vibrating sample magnetometer also failed to show evidence for titanomagnetites. The remanent magnetization is carried by a thermally unstable mineral that breaks down at ~420°C, probably maghemite. The field- and frequency-dependent relationships are also directly opposite to those in the reversal zone, with no signs of titanomagnetite characteristics. Altogether, these rock magnetic data seem to be sensitive indicators of alteration and support the contention that maghemite is responsible for the magnetic signatures displayed in the altered peridotites of the upper section. The magnetic minerals of the basement rocks from Sites 1068, 1069, and 1070 are of variable particle size but fall within the pseudo-single-domain size range (0.2-14 µm). The average natural remanent magnetization (NRM) intensity of recovered serpenitinized peridotite is typically on the order of 20 mA/m for samples from Site 1068, but ~120 mA/m for samples from Site 1070. The much stronger magnetization intensity of Site 1070 is apparently in excellent agreement with the observed magnetic anomaly high. Nearly half of the NRM intensity remained after 400°C demagnetization, suggesting that the remanence can contribute significantly to the marine magnetic anomaly.

Rock magnetic properties of ODP Leg 134 sites

During Leg 134, the influence of ridge collision and subduction on the structural evolution of island arcs was investigated by drilling at a series of sites in the collision zone between the d'Entrecasteaux Zone (DEZ) and the central New Hebrides Island Arc. The DEZ is an arcuate Eocene-Oligocene submarine volcanic chain that extends from the northern New Caledonia Ridge to the New Hebrides Trench. High magnetic susceptibilities and intensities of magnetic remanence were measured in volcanic silts, sands, siltstones, and sandstones from collision zone sites. This chapter presents the preliminary results of studies of magnetic mineralogy, magnetic properties, and magnetic fabric of sediments and rocks from Sites 827 through 830 in the collision zone. The dominant carrier of remanence in the highly magnetic sediments and sedimentary rocks in the DEZ is low-titanium titanomagnetite of variable particle size. Changes in rock magnetic properties reflect variations in the abundance and size of titanomagnetite particles, which result from differences in volcanogenic contribution and the presence or absence of graded beds. Although the anisotropy of magnetic susceptibility results are difficult to interpret in terms of regional stresses because the cores were azimuthally unoriented, the shapes of the susceptibility ellipsoids provide information about deformation style. The magnetic fabric of most samples is oblate, dominated by foliation, as is the structural fabric. The variability of degree of anisotropy (P) and a factor that measures the shape of the ellipsoid (q) reflect the patchy nature of deformation, at a micrometer scale, that is elucidated by scanning electron microscope analysis. The nature of this patchiness implies that deformation in the shear zones is accomplished primarily by motion along bedding planes, whereas the material within the beds themselves remains relatively undeformed.

Disentangling the impact of seroconversion age and set-point viral load on ART-free HIV survival

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

Variable Step Maximum Power Point Tracking for Photovoltaic Modules with Embedded Temperature Sensing

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

Testing Independence in High Dimensions & Identifiability of Graphical Models

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

Typing mineral deposits using their grades and tonnages in an artificial neural network

A test of the ability of a probabilistic neural network to classify deposits into types on the basis of deposit tonnage and average Cu, Mo, Ag, Au, Zn, and Pb grades is conducted. The purpose is to examine whether this type of system might serve as a basis for integrating geoscience information available in large mineral databases to classify sites by deposit type. Benefits of proper classification of many sites in large regions are relatively rapid identification of terranes permissive for deposit types and recognition of specific sites perhaps worthy of exploring further. Total tonnages and average grades of 1,137 well-explored deposits identified in published grade and tonnage models representing 13 deposit types were used to train and test the network. Tonnages were transformed by logarithms and grades by square roots to reduce effects of skewness. All values were scaled by subtracting the variable's mean and dividing by its standard deviation. Half of the deposits were selected randomly to be used in training the probabilistic neural network and the other half were used for independent testing. Tests were performed with a probabilistic neural network employing a Gaussian kernel and separate sigma weights for each class (type) and each variable (grade or tonnage). Deposit types were selected to challenge the neural network. For many types, tonnages or average grades are significantly different from other types, but individual deposits may plot in the grade and tonnage space of more than one type. Porphyry Cu, porphyry Cu-Au, and porphyry Cu-Mo types have similar tonnages and relatively small differences in grades. Redbed Cu deposits typically have tonnages that could be confused with porphyry Cu deposits, also contain Cu and, in some situations, Ag. Cyprus and kuroko massive sulfide types have about the same tonnages. Cu, Zn, Ag, and Au grades. Polymetallic vein, sedimentary exhalative Zn-Pb, and Zn-Pb skarn types contain many of the same metals. Sediment-hosted Au, Comstock Au-Ag, and low-sulfide Au-quartz vein types are principally Au deposits with differing amounts of Ag. Given the intent to test the neural network under the most difficult conditions, an overall 75% agreement between the experts and the neural network is considered excellent. Among the largestclassification errors are skarn Zn-Pb and Cyprus massive sulfide deposits classed by the neuralnetwork as kuroko massive sulfides—24 and 63% error respectively. Other large errors are the classification of 92% of porphyry Cu-Mo as porphyry Cu deposits. Most of the larger classification errors involve 25 or fewer training deposits, suggesting that some errors might be the result of small sample size. About 91% of the gold deposit types were classed properly and 98% of porphyry Cu deposits were classes as some type of porphyry Cu deposit. An experienced economic geologist would not make many of the classification errors that were made by the neural network because the geologic settings of deposits would be used to reduce errors. In a separate test, the probabilistic neural network correctly classed 93% of 336 deposits in eight deposit types when trained with presence or absence of 58 minerals and six generalized rock types. The overall success rate of the probabilistic neural network when trained on tonnage and average grades would probably be more than 90% with additional information on the presence of a few rock types.