889 resultados para REGRESSION TREES
Resumo:
In this thesis, we consider Bayesian inference on the detection of variance change-point models with scale mixtures of normal (for short SMN) distributions. This class of distributions is symmetric and thick-tailed and includes as special cases: Gaussian, Student-t, contaminated normal, and slash distributions. The proposed models provide greater flexibility to analyze a lot of practical data, which often show heavy-tail and may not satisfy the normal assumption. As to the Bayesian analysis, we specify some prior distributions for the unknown parameters in the variance change-point models with the SMN distributions. Due to the complexity of the joint posterior distribution, we propose an efficient Gibbs-type with Metropolis- Hastings sampling algorithm for posterior Bayesian inference. Thereafter, following the idea of [1], we consider the problems of the single and multiple change-point detections. The performance of the proposed procedures is illustrated and analyzed by simulation studies. A real application to the closing price data of U.S. stock market has been analyzed for illustrative purposes.
Resumo:
This morning Dr. Battle will introduce descriptive statistics and linear regression and how to apply these concepts in mathematical modeling. You will also learn how to use a spreadsheet to help with statistical analysis and to create graphs.
Resumo:
Traditional courses and textbooks in occupational safety emphasize rules, standards, and guidelines. This paper describes the early stage of a project to upgrade a traditional college course on fire protection by incorporating learning materials to develop the higher-level cognitive ability known as synthesis. Students will be challenged to synthesize textbook information into fault tree diagrams. The paper explains the place of synthesis in Bloom’s taxonomy of cognitive abilities and the utility of fault trees diagrams as a tool for synthesis. The intended benefits for students are: improved abilities to synthesize, a deeper understanding of fire protection practices, ability to construct fault trees for a wide range of undesired occurrences, and perhaps recognition that heavy reliance on memorization is the hard way to learn occupational safety and health.
Resumo:
OBJECTIVES: This paper is concerned with checking goodness-of-fit of binary logistic regression models. For the practitioners of data analysis, the broad classes of procedures for checking goodness-of-fit available in the literature are described. The challenges of model checking in the context of binary logistic regression are reviewed. As a viable solution, a simple graphical procedure for checking goodness-of-fit is proposed. METHODS: The graphical procedure proposed relies on pieces of information available from any logistic analysis; the focus is on combining and presenting these in an informative way. RESULTS: The information gained using this approach is presented with three examples. In the discussion, the proposed method is put into context and compared with other graphical procedures for checking goodness-of-fit of binary logistic models available in the literature. CONCLUSION: A simple graphical method can significantly improve the understanding of any logistic regression analysis and help to prevent faulty conclusions.
Resumo:
OBJECTIVE: To analyse risk factors in alpine skiing. DESIGN: A controlled multicentre survey of injured and non-injured alpine skiers. SETTING: One tertiary and two secondary trauma centres in Bern, Switzerland. PATIENTS AND METHODS: All injured skiers admitted from November 2007 to April 2008 were analysed using a completed questionnaire incorporating 15 parameters. The same questionnaire was distributed to non-injured controls. Multiple logistic regression was performed. Patterns of combined risk factors were calculated by inference trees. A total of 782 patients and 496 controls were interviewed. RESULTS: Parameters that were significant for the patients were: high readiness for risk (p = 0.0365, OR 1.84, 95% CI 1.04 to 3.27); low readiness for speed (p = 0.0008, OR 0.29, 95% CI 0.14 to 0.60); no aggressive behaviour on slopes (p<0.0001, OR 0.19, 95% CI 0.09 to 0.37); new skiing equipment (p = 0.0228, OR 59, 95% CI 0.37 to 0.93); warm-up performed (p = 0.0015, OR 1.79, 95% CI 1.25 to 2.57); old snow compared with fresh snow (p = 0.0155, OR 0.31, 95% CI 0.12 to 0.80); old snow compared with artificial snow (p = 0.0037, OR 0.21, 95% CI 0.07 to 0.60); powder snow compared with slushy snow (p = 0.0035, OR 0.25, 95% CI 0.10 to 0.63); drug consumption (p = 0.0044, OR 5.92, 95% CI 1.74 to 20.11); and alcohol abstinence (p<0.0001, OR 0.14, 95% CI 0.05 to 0.34). Three groups at risk were detected: (1) warm-up 3-12 min, visual analogue scale (VAS)(speed) >4 and bad weather/visibility; (2) VAS(speed) 4-7, icy slopes and not wearing a helmet; (3) warm-up >12 min and new skiing equipment. CONCLUSIONS: Low speed, high readiness for risk, new skiing equipment, old and powder snow, and drug consumption are significant risk factors when skiing. Future work should aim to identify more precisely specific groups at risk and develop recommendations--for example, a snow weather index at valley stations.
Resumo:
Ultrasonic acoustic emission (UAE) in trees is often related to collapsing water columns in the flow path as a result of tensions that are too strong (cavitation). However, in a decibel (dB) range below that associated with cavitation, a close relationship was found between UAE intensities and stem radius changes. • UAE was continuously recorded on the stems of mature field-grown trees of Scots pine (Pinus sylvestris) and pubescent oak (Quercus pubescens) at a dry inner-Alpine site in Switzerland over two seasons. The averaged 20-Hz records were related to microclimatic conditions in air and soil, sap-flow rates and stem-radius fluctuations de-trended for growth (ΔW). • Within a low-dB range (27 ± 1 dB), UAE regularly increased and decreased in a diurnal rhythm in parallel with ΔW on cloudy days and at night. These low-dB emissions were interrupted by UAE abruptly switching between the low-dB range and a high-dB range (36 ± 1 dB) on clear, sunny days, corresponding to the widely supported interpretation of UAE as sound from cavitations. • It is hypothesized that the low-dB signals in drought-stressed trees are caused by respiration and/or cambial growth as these physiological activities are tissue water-content dependent and have been shown to produce courses of CO2 efflux similar to our courses of ΔW and low-dB UAE.
Resumo:
A combinatorial protocol (CP) is introduced here to interface it with the multiple linear regression (MLR) for variable selection. The efficiency of CP-MLR is primarily based on the restriction of entry of correlated variables to the model development stage. It has been used for the analysis of Selwood et al data set [16], and the obtained models are compared with those reported from GFA [8] and MUSEUM [9] approaches. For this data set CP-MLR could identify three highly independent models (27, 28 and 31) with Q2 value in the range of 0.632-0.518. Also, these models are divergent and unique. Even though, the present study does not share any models with GFA [8], and MUSEUM [9] results, there are several descriptors common to all these studies, including the present one. Also a simulation is carried out on the same data set to explain the model formation in CP-MLR. The results demonstrate that the proposed method should be able to offer solutions to data sets with 50 to 60 descriptors in reasonable time frame. By carefully selecting the inter-parameter correlation cutoff values in CP-MLR one can identify divergent models and handle data sets larger than the present one without involving excessive computer time.
Resumo:
Question: Is stomatal regulation specific for climate and tree species, and does it reveal species-specific responses to drought? Is there a link to vegetation dynamics? Location: Dry inner alpine valley, Switzerland Methods: Stomatal aperture (θE) of Pinus sylvestris, Quercus pubescens, Juniperus communis and Picea abies were continuously estimated by the ratio of measured branch sap flow rates to potential transpiration rates (adapted Penman-Monteith single leaf approach) at 10-min intervals over four seasons. Results: θE proved to be specific for climate and species and revealed distinctly different drought responses: Pinus stomata close disproportionately more than neighbouring species under dry conditions, but has a higher θE than the other species when weather was relatively wet and cool. Quercus keeps stomata more open under drought stress but has a lower θE under humid conditions. Juniperus was most drought-tolerant, whereas Picea stomata close almost completely during summer. Conclusions: The distinct microclimatic preferences of the four tree species in terms of θE strongly suggest that climate (change) is altering tree physiological performances and thus species-specific competitiveness. Picea and Pinus currently live at the physiological limit of their ability to withstand increasing temperature and drought intensities at the sites investigated, whereas Quercus and Juniperus perform distinctly better. This corresponds, at least partially, with regional vegetation dynamics: Pinus has strongly declined, whereas Quercus has significantly increased in abundance in the past 30 years. We conclude that θE provides an indication of a species' ability to cope with current and predicted climate.
Trees, trust and the state: A comparison of participatory forest management in Pakistan and Tanzania
Resumo:
This paper introduces and analyzes a stochastic search method for parameter estimation in linear regression models in the spirit of Beran and Millar [Ann. Statist. 15(3) (1987) 1131–1154]. The idea is to generate a random finite subset of a parameter space which will automatically contain points which are very close to an unknown true parameter. The motivation for this procedure comes from recent work of Dümbgen et al. [Ann. Statist. 39(2) (2011) 702–730] on regression models with log-concave error distributions.