Stratification of the severity of critically ill patients with classification trees

Background: Development of three classification trees (CT) based on the CART (Classification and Regression Trees), CHAID (Chi-Square Automatic Interaction Detection) and C4.5 methodologies for the calculation of probability of hospital mortality; the comparison of the results with the APACHE II, SAPS II and MPM II-24 scores, and with a model based on multiple logistic regression (LR). Methods: Retrospective study of 2864 patients. Random partition (70:30) into a Development Set (DS) n = 1808 and Validation Set (VS) n = 808. Their properties of discrimination are compared with the ROC curve (AUC CI 95%), Percent of correct classification (PCC CI 95%); and the calibration with the Calibration Curve and the Standardized Mortality Ratio (SMR CI 95%). Results: CTs are produced with a different selection of variables and decision rules: CART (5 variables and 8 decision rules), CHAID (7 variables and 15 rules) and C4.5 (6 variables and 10 rules). The common variables were: inotropic therapy, Glasgow, age, (A-a)O2 gradient and antecedent of chronic illness. In VS: all the models achieved acceptable discrimination with AUC above 0.7. CT: CART (0.75(0.71-0.81)), CHAID (0.76(0.72-0.79)) and C4.5 (0.76(0.73-0.80)). PCC: CART (72(69- 75)), CHAID (72(69-75)) and C4.5 (76(73-79)). Calibration (SMR) better in the CT: CART (1.04(0.95-1.31)), CHAID (1.06(0.97-1.15) and C4.5 (1.08(0.98-1.16)). Conclusion: With different methodologies of CTs, trees are generated with different selection of variables and decision rules. The CTs are easy to interpret, and they stratify the risk of hospital mortality. The CTs should be taken into account for the classification of the prognosis of critically ill patients.

Relación entre variables de personalidad e interés por temas morbosos y sexuales en los mass media: réplica del estudio de Zuckerman y Litle en una muestra española

Este trabajo pretende replicar con muestras españolas un estudio americano diseñado para evaluar las relaciones entre el rasgo Búsquesa de Sensaciones, Extraversión, Neuroticismo y Psicoticismo, con el interés por la contemplación o lectura de sucesos violentos, morbosos y sexuales en los medios de comunicación (o reales). El estudio se realizó a partir de inventarios autoinformados de curiosidad morbosa, cuestionarios de personalidad y autopuntuaciones de asistencia a películas pornográficas y de terror. Los resultados obtenidos informan de una relación positiva entre las escalas de Búsqueda de Sensaciones (EBS) y las escalas de Curiosidad Morbosa (CTM-CTS). Ello sugiere como hipótesis que los altos buscadores de sensaciones podrían estar interesados en estímulos que aumentan la actividad del sistema central de catecolaminas. Para ello se parte de hallazgos experimentales previos que apuntan que películas morbosas de terror o sexuales aumentan los niveles de las catecolaminas periféricas.

Rendimiento matemático en contextos bilingües: análisis de la incidencia de algunas variables del contexto socio-educativo

El trabajo que se presenta en este artículo se sitúa en el contexto bilingüe de la provincia de Lleida en la que coexisten dos lenguas en contacto, el catalán y el castellano, si bien existe un predominio de uso familiar y escolar de la primera de ellas. En este contexto, nuestro estudio analiza cómo influyen en el rendimiento en matemáticas las variables del contexto educativo de condición lingüística familiar y situación socio-profesional de la familia, así como las variables individuales de conciente intelectual y autoconcepto matemático. Los resultados globales muestran, en primer lugar, la importancia del lenguaje como instrumento mediador de la enseñanza-aprendizaje de contenidos matemáticos, ya que el rendimiento de los alumnos en esta área curricular está en función del dominio y el uso de la lengua vehicular de la enseñanza —el catalán—. En segundo lugar, se constata una correlación significativa entre el rendimiento matemático y las variables individuales de cociente intelectual y el autoconcepto sobre el propio rendimiento matemático.

Scaling behavior of random knots.

Using numerical simulations we investigate how overall dimensions of random knots scale with their length. We demonstrate that when closed non-self-avoiding random trajectories are divided into groups consisting of individual knot types, then each such group shows the scaling exponent of approximately 0.588 that is typical for self-avoiding walks. However, when all generated knots are grouped together, their scaling exponent becomes equal to 0.5 (as in non-self-avoiding random walks). We explain here this apparent paradox. We introduce the notion of the equilibrium length of individual types of knots and show its correlation with the length of ideal geometric representations of knots. We also demonstrate that overall dimensions of random knots with a given chain length follow the same order as dimensions of ideal geometric representations of knots.

Bond strength tests of dental adhesive systems and their correlation with clinical results : a meta-analysis

OBJECTIVE: To evaluate the variability of bond strength test results of adhesive systems (AS) and to correlate the results with clinical parameters of clinical studies investigating cervical restorations. MATERIALS AND METHODS: Regarding the clinical studies, the internal database which had previously been used for a meta-analysis on cervical restorations was updated with clinical studies published between 2008 and 2012 by searching the PubMed and SCOPUS databases. PubMed and the International Association for Dental Research abstracts online were searched for laboratory studies on microtensile, macrotensile and macroshear bond strength tests. The inclusion criteria were (1) dentin, (2) testing of at least four adhesive systems, (3) same diameter of composite and (4) 24h of water storage prior to testing. The clinical outcome variables were retention loss, marginal discoloration, detectable margins, and a clinical index comprising the three parameters by weighing them. Linear mixed models which included a random study effect were calculated for both, the laboratory and the clinical studies. The variability was assessed by calculating a ratio of variances, dividing the variance among the estimated bonding effects obtained in the linear mixed models by the sum of all variance components estimated in these models. RESULTS: Thirty-two laboratory studies fulfilled the inclusion criteria comprising 183 experiments. Of those, 86 used the microtensile test evaluating 22 adhesive systems (AS). Twenty-seven used the macrotensile test with 17 AS, and 70 used the macroshear test with 24 AS. For 28 AS the results from clinical studies were available. Microtensile and macrotensile (Spearman rho=0.66, p=0.007) were moderately correlated and also microtensile and macroshear (Spearman rho=0.51, p=0.03) but not macroshear and macrotensile (Spearman rho=0.34, p=0.22). The effect of the adhesive system was significant for microtensile and macroshear (p<0.001) but not for macrotensile. The effect of the adhesive system could explain 36% of the variability of the microtensile test, 27% of the macrotensile and 33% of the macroshear test. For the clinical trials, about 49% of the variability of retained restorations could be explained by the adhesive system. With respect to the correlation between bond strength tests and clinical parameters, only a moderate correlation between micro- and macrotensile test results and marginal discoloration was demonstrated. However, no correlation between these tests and a retention loss or marginal integrity was shown. The correlation improved when more studies were included compared to assessing only one study. SIGNIFICANCE: The high variability of bond strength test results highlights the need to establish individual acceptance levels for a given test institute. The weak correlation of bond-strength test results with clinical parameters leads to the conclusion that one should not rely solely on bond strength tests to predict the clinical performance of an adhesive system but one should conduct other laboratory tests like tests on the marginal adaptation of fillings in extracted teeth and the retention loss of restorations in non-retentive cavities after artificial aging.

Viticultura de precisión: predicción de cosecha a partir de variables del cultivo e índices de vegetación

El presente trabajo muestra los resultados provisionales de un proyecto de viticultura de precisión, llevado a cabo entre la Universidad de Lleida y la empresa Codorníu, S.A., con el objeto de predecir el rendimiento, de forma localizada, de diferentes variedades de viña para la producción de vinos y cavas de calidad. Las variables de cultivo muestreadas de forma detallada han sido: número de yemas, número de brotes, número de racimos, peso de la poda, vigor del cultivo y cosecha del año anterior (2003). El vigor del cultivo se ha extraído de una imagen multiespectral Quickbird 2. Estas variables se han analizado en un modelo de regresión múltiple con la cosecha del año 2004 (variable dependiente), ofreciendo diferentes ajustes según la variedad considerada y variables incorporadas en el análisis.

The economic relevance of climate variables in agriculture: the case of Spain

We estimate how climate variables affect price and acreage of productive farmland using the Ricardian approach. Furthermore, we use our estimations to evaluate the joint effects of possible cli- mate changes within the time horizon of 2010 and 2050. Our results show that the price of rainfed land in Spain tends to increase but rainfed acreage decreases. On the other hand, the effect on irrigated farmland price and acreage presents some mixed results, however, in the long run the dominant pattern is clearly increasing for both prices and acreage.

Generating highly balanced sudoku problems as hard problems

Sudoku problems are some of the most known and enjoyed pastimes, with a never diminishing popularity, but, for the last few years those problems have gone from an entertainment to an interesting research area, a twofold interesting area, in fact. On the one side Sudoku problems, being a variant of Gerechte Designs and Latin Squares, are being actively used for experimental design, as in [8, 44, 39, 9]. On the other hand, Sudoku problems, as simple as they seem, are really hard structured combinatorial search problems, and thanks to their characteristics and behavior, they can be used as benchmark problems for refining and testing solving algorithms and approaches. Also, thanks to their high inner structure, their study can contribute more than studies of random problems to our goal of solving real-world problems and applications and understanding problem characteristics that make them hard to solve. In this work we use two techniques for solving and modeling Sudoku problems, namely, Constraint Satisfaction Problem (CSP) and Satisfiability Problem (SAT) approaches. To this effect we define the Generalized Sudoku Problem (GSP), where regions can be of rectangular shape, problems can be of any order, and solution existence is not guaranteed. With respect to the worst-case complexity, we prove that GSP with block regions of m rows and n columns with m = n is NP-complete. For studying the empirical hardness of GSP, we define a series of instance generators, that differ in the balancing level they guarantee between the constraints of the problem, by finely controlling how the holes are distributed in the cells of the GSP. Experimentally, we show that the more balanced are the constraints, the higher the complexity of solving the GSP instances, and that GSP is harder than the Quasigroup Completion Problem (QCP), a problem generalized by GSP. Finally, we provide a study of the correlation between backbone variables – variables with the same value in all the solutions of an instance– and hardness of GSP.

The impact of balancing on problem hardness in a highly structured domain

Random problem distributions have played a key role in the study and design of algorithms for constraint satisfaction and Boolean satisfiability, as well as in ourunderstanding of problem hardness, beyond standard worst-case complexity. We consider random problem distributions from a highly structured problem domain that generalizes the Quasigroup Completion problem (QCP) and Quasigroup with Holes (QWH), a widely used domain that captures the structure underlying a range of real-world applications. Our problem domain is also a generalization of the well-known Sudoku puz- zle: we consider Sudoku instances of arbitrary order, with the additional generalization that the block regions can have rectangular shape, in addition to the standard square shape. We evaluate the computational hardness of Generalized Sudoku instances, for different parameter settings. Our experimental hardness results show that we can generate instances that are considerably harder than QCP/QWH instances of the same size. More interestingly, we show the impact of different balancing strategies on problem hardness. We also provide insights into backbone variables in Generalized Sudoku instances and how they correlate to problem hardness.

The economic relevance of climate variables in agriculture: The case of Spain

We estimate the impact of the different climate variables on the value of Spanish farmland. We distinguish between irrigated and non-irrigated lands and use data on temperature, precipitations, physical and socioeconomic characteristics to measure these effects on farm prices and in the number of hectares of farmland. We conclude presenting the main results of our analysis, region by region, and examining the policies that could be more effective to prevent undesired effects.