1,720,977 research outputs found
Probabilistic forecast reconciliation with applications to wind power and electric load
Get PDFNew methods are proposed for adjusting probabilistic forecasts to ensure coherence with the aggregation constraints inherent in temporal hierarchies. The different approaches nested within this framework include methods that exploit information at all levels of the hierarchy as well as a novel method based on cross-validation. The methods are evaluated using real data from two wind farms in Crete and electric load in Boston. For these applications, optimal decisions related to grid operations and bidding strategies are based on coherent probabilistic forecasts of energy power. Empirical evidence is also presented showing that probabilistic forecast reconciliation improves the accuracy of the probabilistic forecasts
Bayesian density forecasting of intraday electricity prices using multivariate skew t distributions
No full textElectricity spot prices exhibit strong time series properties, including substantial periodicity, both inter-day and intraday serial correlation, heavy tails and skewness. In this paper we capture these characteristics using a first order vector autoregressive model with exogenous effects and a skew t distributed disturbance. The vector is longitudinal, in that it comprises observations on the spot price at intervals during a day. A band two inverse scale matrix is employed for the disturbance, as well as a sparse autoregressive coefficient matrix. This corresponds to a parsimonious dependency structure that directly relates an observation to the two immediately prior, and the observation at the same time the previous day. We estimate the model using Markov Chain Monte Carlo, which allows for the evaluation of the complete predictive distribution of future spot prices. We apply the model to hourly Australian electricity spot prices observed over a three year period, with four different nested multivariate error distributions: skew t, symmetric t, skew normal and symmetric normal. The forecasting performance is judged over a 30Â day forecast trial using the continuous ranked probability score, which accounts for both predictive bias and sharpness.C11 C13 C53 Vector Autoregression Longitudinal Model Parsimonious Covariance Asymmetry Continuous Ranked Probability Score Electricity Spot Price Forecasting
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Bayesian identification, selection and estimation of semiparametric functions in high-dimensional additive models
No full textIn this paper we propose an approach to both estimate and select unknown smooth functions in an additive model with potentially many functions. Each function is written as a linear combination of basis terms, with coefficients regularized by a proper linearly constrained Gaussian prior. Given any potentially rank deficient prior precision matrix, we show how to derive linear constraints so that the corresponding effect is identified in the additive model. This allows for the use of a wide range of bases and precision matrices in priors for regularization. By introducing indicator variables, each constrained Gaussian prior is augmented with a point mass at zero, thus allowing for function selection. Posterior inference is calculated using Markov chain Monte Carlo and the smoothness in the functions is both the result of shrinkage through the constrained Gaussian prior and model averaging. We show how using non-degenerate priors on the shrinkage parameters enables the application of substantially more computationally efficient sampling schemes than would otherwise be the case. We show the favourable performance of our approach when compared to two contemporary alternative Bayesian methods. To highlight the potential of our approach in high-dimensional settings we apply it to estimate two large seemingly unrelated regression models for intra-day electricity load. Both models feature a variety of different univariate and bivariate functions which require different levels of smoothing, and where component selection is meaningful. Priors for the error disturbance covariances are selected carefully and the empirical results provide a substantive contribution to the electricity load modelling literature in their own right.
Statistische Modellierung mit finiten Skew-t Mixture Verteilungen und deren Anwendungen in den Life Sciences
No full textThere is a need for flexible distribution and error models considering multivariate data in the life sciences. This thesis explores a flexible class of multivariate models, especially suited for data encountered in real-life situations, where outliers and skewness are prevalent and provides an algorithm for fitting such models to use them efficiently in practice.
We introduce univariate and multivariate mixtures of skew-t distributions, beginning with their mathematical definitions and followed by their hierarchical representation, which facilitates the implementation of the Expectation-Maximization (EM) algorithm for parameter estimation. A new proposal for classifying multivariate clinical data is proposed, based on fitting multivariate skew-t mixtures separately to patients from diseased case and non-diseased control groups. The ratio of multivariate densities for cases to controls forms a likelihood ratio, which multiplied by the ratio of prior probabilities of being a case versus control, leads to the posterior odds of being a case. The posterior odds of cases are back-transformed to the probability scale, resulting in individual predictions. The form of the density ratio is discussed for different situations, such as for the constraint of equal variance.
We construct efficient EM algorithms that can accommodate collapsed clusters. A collapsed cluster can be viewed as a distribution that places all of its mass on a lower-dimensional space with no variance. Our approach for the applications targeted in this thesis is that the same underlying data generating process applies to collapsed and non-collapsed clusters. We develop and publish a novel R package fitmixst4, which implements the EM algorithm for fitting multivariate mixtures of skew-t distributions and differentiates collapsed clusters from non-collapsed ones by restricting the variance of the latter to be above a specific bound.
We implement the algorithm in two applications. The first application concerns the update of a leading online clinical risk prediction model for prostate cancer on biopsy to incorporate two novel serum markers. We fit the multivariate skew-t mixtures to the bivariate distribution of the two markers in a sample of cancer cases and controls, respectively, thus forming the likelihood ratio. Prior odds of prostate cancer for individual patients are formed based on their standard clinical risk factor profiles from an existing online risk prediction tool. Multiplication by the likelihood ratio leads to updated individualized posterior probabilities of prostate cancer that combines information from standard risk factors with the new markers. We implement the resulting risk tool with the R package shiny and post it online at the Cleveland Clinic Risk Library to make it accessible to patients and clinicians worldwide.
For the second application, we fit mixtures of multivariate skew-t distributions with collapsed clusters to describe and classify trees that experience mortality versus not from a network of European Beech trees. We model up to five individual tree characteristics and competition indices to form risk prediction models for tree mortality. We visualize two-dimensional contour plots of predictive characteristics for trees experiencing versus not experiencing mortality in order to facilitate communication with forest researchers concerning indicators for mortality. Using separate training and test sets, we show that skew-t based methods slightly outperform traditional logistic regression.
This thesis provides means for life science researchers to implement an intricate modeling framework in order to maximize prediction of outcomes, as well as an understanding of underlying complex nonlinear associations among risk factors. The published R package facilitates implementation, bringing the impact of the models to applications in many fields beyond those shown in this thesis.Es gibt erheblichen Bedarf an flexiblen Verteilungen und Fehlermodellen, um multivariate Daten in Life Science zu untersuchen. Die vorliegende Arbeit untersucht eine flexible Klasse von multivariaten Modellen, die sich besonders gut für reale Daten eignet, in welchen Ausreißer und Schiefen vorhanden sind. Dazu stellt sie einen Algorithmus für die Anpassung solcher Modelle zur Verfügung.
Wir führen zunächst univariate und multivariate Mixture Skew-t Verteilungen ein, beginnend mit deren mathematischer Definition, gefolgt von deren hierarchischer Darstellung, die die Implementation des Expectation-Maximisation (EM) Algorithmus für die Parameterschätzung ermöglicht. Eine neue Interpretation für die Klassifizierung von multivariaten Daten, basierend auf der Anpassung separater multivariater Skew-t Mixtures jeweils für Fall- und Kontrollgruppe, wird vorgeschlagen. Der Quotient von multivariaten Dichten für Fall- und Kontrollgruppe formt eine Likelihood Ratio, die multipliziert mit der priori Wahrscheinlichkeit zu posteriori Odds führt. Die posteriori Odds werden auf die Wahrscheinlichkeitsskala zurücktransformiert. Die Form der Dichtequotienten wird für unterschiedliche Situationen, wie etwa für den Fall von gleichen Varianzen, diskutiert.
Wir konstruieren einen effizienten EM Algorithmus, der mit Collapsed Cluster umgehen kann. Ein Collapsed Cluster kann man als eine Verteilung betrachten, das alle Masse in einem niedrigeren dimensionalen Raum ohne Varianz hat. Unser Ansatz für die Applikationen in dieser Arbeit ist die Annahme, dass der Prozess, der die Daten generiert, für Collapsed und Non-Collapsed Clusters derselbe ist. Des weiteren entwickeln und veröffentlichen wir ein neues R Paket fitmixst4, das den EM Algorithmus für die Anpassung von multivariaten Skew-t Mixtures Verteilungen implementiert und Collapsed Clusters von normalen Gruppen differenziert.
Wir verwenden den Algorithmus in zwei Anwendungen. Die erste Anwendung behandelt das Update eines führenden klinischen Online-Risikoprädiktionsmodel für Prostatakrebs mit Biopsien, in welches zwei neue Serummarker eingebaut werden. Wir schätzen multivariate Skew-t Mixtures für die bivariate Verteilung der beiden Marker für Krebs- und Kontrollfälle, um eine Likelihood Ratio zu bekommen. Die priori Odds für Prostatakrebs für individuelle Patienten werden basierend auf klinischen Standardrisikofaktorprofilen mit dem existierenden Online-Risikoprädiktionstool berechnet. Die Multiplikation mit der Likelihood Ratio führt zu angepassten individualisierten posteriori Wahrscheinlichkeiten für Prostatakrebs, die die Information der Standardrisikofaktoren mit den neuen Markern kombiniert. Wir implementieren das resultierende Risikotool mit dem R Paket shiny und stellen es online auf der Cleveland Clinic Risk Library zur Verfügung, um es weltweit für Patienten und Kliniker zugänglich zu machen.
Für die zweite Anwendung haben wir Multivariate Skew-t Mixtures mit Collapsed Clustern verwendet, die die Sterbewahrscheinlichkeit von Bäumen in einem europäischen Netzwerk für Buchen beschreiben und klassifizieren. Wir modellieren bis zu fünf individuelle Baumcharakteristiken und Wettbewerbsindizes, um ein Risikoprädiktionsmodel für die Sterblichkeit der Bäume zu entwickeln. Zusätzlich haben wir zweidimensionale Konturdiagramme der prädiktiven Charakteristiken visualisiert, um eine Grundlage für die Kommunikation mit Forstwissenschaftlern zu schaffen. Mit Hilfe von separaten Trainings- und Validierungssets, kann gezeigt werden, dass der Ansatz mit den Skew-t Verteilungen die traditionelle logistische Regression übertrifft.
Die vorliegende Arbeit stellt Forschern in den Life Sciences ein komplexes Modeling Framework zur Verfügung, welches die Prädiktionsresultate maximiert und das Verständnis der zu Grunde liegenden nicht-linearen Assoziationen veranschaulicht. Das publizierte R Paket erleichtert die Implementation, um die Anwendbarkeit dieser Modelle auf andere Sachgebiete zu übertragen
Bayesian skew selection for multivariate models
No full textWe develop a Bayesian approach for the selection of skew in multivariate skew t distributions constructed through hidden conditioning in the manners suggested by either Azzalini and Capitanio (2003) or Sahu et al. (2003). We show that the skew coefficients for each margin are the same for the standardized versions of both distributions. We introduce binary indicators to denote whether there is symmetry, or skew, in each dimension. We adopt a proper beta prior on each non-zero skew coefficient, and derive the corresponding prior on the skew parameters. In both distributions we show that as the degrees of freedom increases, the prior smoothly bounds the non-zero skew parameters away from zero and identifies the posterior. We estimate the model using Markov chain Monte Carlo (MCMC) methods by exploiting the conditionally Gaussian representation of the skew t distributions. This allows for the search through the posterior space of all possible combinations of skew and symmetry in each dimension. We show that the proposed method works well in a simulation setting, and employ it in two multivariate econometric examples. The first involves the modeling of foreign exchange rates and the second is a vector autoregression for intra-day electricity spot prices. The approach selects skew along the original coordinates of the data, which proves insightful in both examples.Skew t distribution MCMC Point priors Model averaging Vector autoregression Intra-day electricity prices
Variations on the Author
Get PDF“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Get PDFWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Dispelling the Myths Behind First-author Citation Counts
Get PDFWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
- …
