stats@lmu @statsatlmu - Tumblr Blog

Department of Statistics joins LMU Open Science Center

Our department is now a member of the LMU Open Science Center.

By joining the OSC, we express that we actively support the goals of the OSC.

We want to take on a leading role in fostering good research practices and promoting open science. We believe that statistics as a subject plays a major role in solving the reproducibility crisis and in working towards better, more open research.

More about the LMU OSC:

Website: https://www.osc.uni-muenchen.de

Twitter: @lmu_osc

KOALA: a new paradigm for election coverage

Just published in AStA Advances in Statistical Analysis by Alexander Bauer, Andreas Bender, André Klima & Helmut Küchenhoff, the paper presents "an opinion poll-based 'now-cast' of probabilities of events in multi-party electoral systems."

Abstract:

Common election poll reporting is often misleading as sample uncertainty is addressed insufficiently or not covered at all. Furthermore, main interest usually lies beyond the simple party shares. For a more comprehensive opinion poll and election coverage, we propose shifting the focus toward the reporting of survey-based probabilities for specific events of interest. We present such an approach for multi-party electoral systems, focusing on probabilities of coalition majorities. A Monte Carlo approach based on a Bayesian Multinomial-Dirichlet model is used for estimation. Probabilities are estimated, assuming the election was held today (“now-cast”), not accounting for potential shifts in the electorate until election day (“fore-cast”). Since our method is based on the posterior distribution of party shares, the approach can be used to answer a variety of questions related to the outcome of an election. We also introduce visualization techniques that facilitate a more adequate depiction of relevant quantities as well as respective uncertainties. The benefits of our approach are discussed by application to the German federal elections in 2013 and 2017. An open-source implementation of our methods is freely available in the R package coalitions.

Full paper

Congratulations on another great effort by this all-star team of StabLab-Alumni.

#publication

Just published: “Simplified vine copula models: Approximations based on the simplifying assumption.”

Spanhel, Fabian and Malte S. Kurz, "Simplified vine copula models: Approximations based on the simplifying assumption", Electronic Journal of Statistics 13 (2019), no. 1, 1254--1291. doi:10.1214/19-EJS1547.

Abstract:

Vine copulas, or pair-copula constructions, have become an important tool in high-dimensional dependence modeling. Commonly, it is assumed that the data generating copula can be represented by a simplified vine copula (SVC). In this paper, we study the simplifying assumption and investigate the approximation of multivariate copulas by SVCs. We introduce the partial vine copula (PVC) which is a particular SVC where to any edge a j-th order partial copula is assigned. The PVC generalizes the partial correlation matrix and plays a major role in the approximation of copulas by SVCs. We investigate to what extent the PVC describes the dependence structure of the underlying copula. We show that, in general, the PVC does not minimize the Kullback-Leibler divergence from the true copula if the simplifying assumption does not hold. However, under regularity conditions, stepwise estimators of pair-copula constructions converge to the PVC irrespective of whether the simplifying assumption holds or not. Moreover, we elucidate why the PVC is often the best feasible SVC approximation in practice.

Congrats, Fabian & Malte!

#publication

MünchenerUniMagazin widmet sich dem Thema Künstliche Intelligenz und Data Science

In der aktuellen Ausgabe des MünchnerUniMagazins werden die verschiedenen Aspekte der künstlichen Intelligenz an der LMU vorgestellt. Dabei wird sowohl der Studiengang Data Science als auch die jüngst gestartete Kooperation mit der Tel Aviv University vorgestellt.

StaBLab receives 3-year funding for TourIST

Our StaBLab has received a 3-year grant from the German Research Foundation (DFG) for the interdisciplinary research project TourIST - Tourism In Space and Time.

The aim of the research project is to investigate touristic behavior and its changes in a spatio-temporal context and to determine influencing factors and triggers of the changes in order to be able to better estimate and predict future developments. The data basis are nationwide representative surveys of the "Reiseanalyse", which are conducted annually by Forschungsgemeinschaft Urlaub und Reisen e.V. (FUR). Data from 1995 to 2019 will be included.

The interdisciplinary research project TourIST was conceived in cooperation with Prof. Dr. Jürgen Schmude and Dr. Marion Karl from the Department of Geography of the LMU Munich.

Universitätsstudie Bayernwahl USBW18 (München - Passau - Regensburg)

Aus unserem Lehrforschungsprojekt USBW18 zur Landtagswahl am 14.10.2018 liegen die ersten vorläufigen Ergebnisse vor.

Die Grafik zeigt vorläufige Ergebnisse der Wählerwanderung in Bayern von der Bundestagswahl 2017 zur Landtagswahl 2018 (saldierte Wanderungsströme, Grafik: Statistisches Beratungslabor StaBLab der LMU)

Weitere Ergebnisse und Informationen finden Sie hier.

Best Poster Award in Warsaw!!!

Sevag Kevork was awarded the Best Poster Award at the COSTNET Conference in Warsaw. He was able to convince the local jury with outstanding layout, quality and the scientific content of his poster as well as his striking arguments in the discussion!

We congratualte and are proud of Sevag!

#Awards #highpotentials #poster #network

Hyuuge success at IWSM 2018!

Two awards at the 33rd International Workshop on Statistical Modelling held in Bristol this year went to LMU statistics researchers and their teams:

Michael Lebacher won the Best Student Talk award for his presentation International Arms Transfers: A Dynamic Separable Network Model With Heterogeneity Components (joint work with Paul W. Thurner and Göran Kauermann)

Alexander Bauer won the coveted Best Student Poster award for his poster KOALA: A new paradigm for election coverage in multi-party electoral systems. (joint work with Andreas Bender, André Klima and Helmut Küchenhoﬀ, also check out their snazzy website)

Congratulations to all those involved, we expect there will be champagne & cake in the Alte Bib upon your return!

#awards highpotentials werollintheyhatin

Just published: “Anticipation of food safety and fraud issues: ISAR - A new screening tool to monitor food prices and commodity flows.”

by K. Verhaelen, A. Bauer, F. Günther, B. Müller, M. Nist, B. Ülker Celik, C. Weidner, H. Küchenhoff, P.Wallner, in Food Control

Abstract:

Despite major efforts, food safety and food fraud issues continue to appear, showing the need for an improved risk anticipation to enhance consumer protection. To this end, we need a holistic view on the food production beyond the classical farm to fork principle and consider developments in- and outside the food chain. Commodity flows and prices are considered as relevant drivers for the emergence of food safety and fraud issues. To conveniently obtain information on these drivers and to monitor their changes, we developed a tool to analyze the volume, origin and price of foods imported to Germany, based on data of the German Foreign Trade Statistic. The ISAR-Tool (Import Screening for the Anticipation of Food Risks) facilitates a descriptive analysis of the import data of about 2400 different food items imported from 220 countries and an automated analysis of the data for a systematic detection of unexpected changes in import volumes and prices. For the automated approach, time series analyses using SARIMA and ETS models are applied and data points outside the model-based 99% prediction intervals are considered as potentially relevant changes. Monthly, more than 2000 of such changes are detected, leaving us with the challenging task of their evaluation for food safety and food fraud. To deal with the vast amount of results, of which the great majority cannot be linked to food safety or fraud issues, we implemented prioritization routines, and the 200 most critical results are considered for further expert evaluation. Despite remaining challenges in the performed evaluation process of the automated analysis, the described approach facilitates already at this stage an improved identification of food safety and food fraud hazards and the implementation of more targeted food safety controls, as shown in this paper. Our most prominent example of an identified food fraud and food safety risk was the detection of massive adulterations of hazelnut products with almonds and cashews, which were associated to the observed all-time high in hazelnut prices. Based on our findings, we conclude that the ISAR-Tool and the described analyses of commodity flows and prices can be important elements in the anticipation of risks and enable an enhanced consumer protection.

https://doi.org/10.1016/j.foodcont.2018.06.029; free access

Congratulations to the StabLab Team!

#publication nuts

Our very own Heidi Seibold is Curator of the Real Scientists Twitter Account @realsci_DE this week!

Just a heads-up:

This week Heidi Seibold curates the twitter account @realsci_DE, the german language Real Scientists Twitter account. @realsci_DE is a rotational Twitter account featuring scientists, researchers, clinicians, writers, communicators and policy makers talking about their lives and their work for a week.

If you want to learn about the life and work of LMU statistician Heidi Seibold (@HeidiBaya) visit https://twitter.com/realsci_DE this week.

#fame twitterati dittschitällnäitiff

Just published: “Bootstrap inference when using multiple imputation”

M. Schomaker, C. Heumann: “Bootstrap inference when using multiple imputation”, Statistics in Medicine

Abstract:

Many modern estimators require bootstrapping to calculate confidence intervals because either no analytic standard error is available or the distribution of the parameter of interest is nonsymmetric. It remains however unclear how to obtain valid bootstrap inference when dealing with multiple imputation to address missing data. We present 4 methods that are intuitively appealing, easy to implement, and combine bootstrap estimation with multiple imputation. We show that 3 of the 4 approaches yield valid inference, but that the performance of the methods varies with respect to the number of imputed data sets and the extent of missingness. Simulation studies reveal the behavior of our approaches in finite samples. A topical analysis from HIV treatment research, which determines the optimal timing of antiretroviral treatment initiation in young children, demonstrates the practical implications of the 4 methods in a sophisticated and realistic setting. This analysis suffers from missing data and uses the g‐formula for inference, a method for which no standard errors are available.

Find the full article here.

#publication

DFG Homepage launched. International Trade of Arms: A Network Approach

The joint project of the Chair of Applied Statistics in Social Sciences, Economics and Business, together with the Chair of Empirical Political Research and Policy Research gives a network perspective for the empirical analysis of the international trade of arms. The identification of flow structures, of their dynamics, and of their causes are a novel contribution to the positive science of peace and conflict. Our quantitatively-oriented network approach is based on measurements of the export, transfer, and import of arms over long time periods (1950-2016). We analyze major conventional weapons (MCW) and small arms and light weapons (SALW).

For more information see our new launched homepage: www.arms-trade-research.uni-muenchen.de

#research #homepage #arms trade

Just published: A note on low-dimensional Kalman smoothers for systems with lagged states in the measurement equation

Abstract:

In this paper we derive a modified Kalman smoother for state space systems with lagged states in the measurement equation. This modified Kalman smoother minimizes the mean squared error (MSE). Computationally efficient algorithms that can be used to implement it in practice are discussed. We also show that the conjecture in Nimark (2015) that the output of his modified Kalman filter for this type of systems can be plugged into the standard Kalman smoother is in general not correct. The competing smoothers are compared with regards to the MSE.

Kurz, M. S. (2018), "A note on low-dimensional Kalman smoothers for systems with lagged states in the measurement equation", Economics Letters 168, pp. 42-45. [full text]

#publication

Just published: Multiple imputation with sequential penalized regression

Just published in Statistical Methods in Medical Research:

Multiple imputation with sequential penalized regression by Faisal M. Zahid and Christian Heumann.

Abstract:

Missing data is a common issue that can cause problems in estimation and inference in biomedical, epidemiological and social research. Multiple imputation is an increasingly popular approach for handling missing data. In case of a large number of covariates with missing data, existing multiple imputation software packages may not work properly and often produce errors. We propose a multiple imputation algorithm called mispr based on sequential penalized regression models. Each variable with missing values is assumed to have a different distributional form and is imputed with its own imputation model using the ridge penalty. In the case of a large number of predictors with respect to the sample size, the use of a quadratic penalty guarantees unique estimates for the parameters and leads to better predictions than the usual Maximum Likelihood Estimation (MLE), with a good compromise between bias and variance. As a result, the proposed algorithm performs well and provides imputed values that are better even for a large number of covariates with small samples. The results are compared with the existing software packages mice, VIM and Amelia in simulation studies. The missing at random mechanism was the main assumption in the simulation study. The imputation performance of the proposed algorithm is evaluated with mean squared imputation error and mean absolute imputation error. The mean squared error, parameter estimates with their standard errors and confidence intervals are also computed to compare the performance in the regression context. The proposed algorithm is observed to be a good competitor to the existing algorithms, with smaller mean squared imputation error, mean absolute imputation error and mean squared error. The algorithm’s performance becomes considerably better than that of the existing algorithms with increasing number of covariates, especially when the number of predictors is close to or even greater than the sample size. Two real-life datasets are also used to examine the performance of the proposed algorithm using simulations.

Find the full paper here.

Neuer Imagefilm zum BA Statistik Studiengang seit heute online (s.o.), auch unter http://www.statistik.uni-muenchen.de/studieninteressierte/index.html

Neues aus dem StaBLab: Dem Lebensmittelbetrug auf der Spur

Eine am StaBLab zusammen mit Fachleuten des Bayerischen Landesamtes für Gesundheit und Lebensmittelsicherheit (LGL) entwickelte Analysemethode „Import screening for the analysis of food risks“ (ISAR) wird eingesetzt um "Gesundheitsrisiken und Betrugspotenzial bei Lebensmitteln so früh wie möglich zu erkennen", so Helmut Küchenhoff. Die interessante Pressemitteilung der LMU dazu findet sich hier.

Just published: “Penalized estimation of complex, non-linear exposure-lag-response associations”

by A. Bender, F. Scheipl, W. Hartl, A. G. Day, H. Küchenhoff

Abstract:

We propose a novel approach for the flexible modeling of complex exposure-lag-response associations in time-to-event data, where multiple past exposures within a defined time window are cumulatively associated with the hazard. Our method allows for the estimation of a wide variety of effects, including potentially smooth and smoothly time-varying effects as well as cumulative effects with leads and lags, taking advantage of the inference methods that have recently been developed for generalized additive mixed models. We apply our method to data from a large observational study of intensive care patients in order to analyze the association of both the timing and the amount of artificial nutrition with the short term survival of critically ill patients. We evaluate the properties of the proposed method by performing extensive simulation studies and provide a systematic comparison with related approaches.

Find the article here.

#publication

Trending Blogs

Recently Viewed Blogs

stats@lmu