ENBIS-13 in Ankara

15 – 19 September 2013 Abstract submission: 5 February – 5 June 2013

My abstracts

 

The following abstracts have been accepted for this event:

  • Strategies for Outlier Detection in Linear Regression with Functional Response

    Authors: André Rehage (TU Dortmund University), Sonja Kuhnt (TU Dortmund University)
    Primary area of focus / application: Modelling
    Keywords: Thermal spraying process, Generalized functional linear models, Outlier detection, Data depth
    Submitted at 29-May-2013 15:31 by André Rehage
    Accepted
    16-Sep-2013 15:25 Strategies for Outlier Detection in Linear Regression with Functional Response
    Functional data analysis is a branch of statistics capturing the specific properties of data which are continuously observed over a certain interval and therefore assumed to be smooth [1]. For instance, functional data occur in meteorology (like temperature over time) or psychometrics (like the ability space curve); moreover the functional aspect is intrinsic in most industrial processes. Recently, statistical methods like principal component analysis or generalized linear models [2] have been extended to functional data. We focus on generalized linear models with functional response and scalar covariates. As in every data analysis also in the analysis of functional data outliers, hence unusual observations can occur. We use the concept of data depth which has been extended to functional data [3] as well as an own approach to define outlier detection methods in the functional case.

    Our work is motivated by an application to a thermal spraying process. Here, particle properties in flight are measured during the whole spraying process, which can be influenced by a number of process parameters. The particle properties can be assumed to be functional response variables in a generalized linear model of the spraying process. This real life example will be scanned for location and shape outliers.

    References:

    [1] Ramsay, J.O. and Silverman, B.W. (2005): Functional Data Analysis, 2nd edition, Springer, New York.
    [2] Müller, H.-G. and Stadtmüller U. (2005): Generalized Functional Linear Models, Ann. Stat., 33(2), 774-805.
    [3] López-Pintado, S. and Romo, J. (2009): On the Concept of Depth for Functional Data, JASA, 104(486), 718-734.
  • Fuzzy Analysis of Health Information System Users’ Security Awareness

    Authors: Ozlem Aydın (Baskent University), Oumout Chouseinoglou (Baskent University)
    Primary area of focus / application: Other: fuzzy
    Keywords: fuzzy sets and fuzzy numbers, fuzzy evaluation, health information systems, health information systems security awareness
    Submitted at 29-May-2013 16:31 by Ozlem Testik
    Accepted (view paper)
    16-Sep-2013 15:25 Fuzzy Analysis of Health Information System Users’ Security Awareness
    Human perceptions, when expressed in natural language, are subjective and vague, and result to the same words indicating different perceptions. Fuzzy sets and fuzzy numbers are efficient tools in describing linguistic variables, especially where personal understandings and perceptions are taken into consideration. Moreover, computer and information security is and always has been fuzzy, and therefore, when assessing security related topics, it is appropriate to use tools designed to deal with fuzziness. In this study, first we have measured the awareness of HIS users via questionnaire, where responses are collected on a 5-point Likert scale. In order to enable the analysis; the linguistic expressions given on the scale are converted to fuzzy numbers, depending on the level of required security in health information systems (HIS). By using fuzzy logic, the subjective judgments have been removed and the linguistic expressions have been modeled accordingly.
  • A Comparison between Minimum Volume Frequentist Confidence Intervals and HPD Bayesian Credibility Intervals

    Authors: Kristina Lurz (University of Würzburg), Rainer Göb (University of Würzburg), Antonio Pievatolo (CNR-IMATI)
    Primary area of focus / application: Other: ISBA/IS Invited Session
    Keywords: Confidence intervals for a probability, HPD credibility intervals, Binomial distribution, Zero-inflated counts, Auditing
    Submitted at 29-May-2013 17:21 by Kristina Lurz
    Accepted
    18-Sep-2013 10:55 A Comparison between Minimum Volume Frequentist Confidence Intervals and HPD Bayesian Credibility Intervals
    Two-sided confidence intervals for the parameter p of a Binomial distribution under a prescribed confidence level are an elementary tool of statistical data analysis. Göb & Lurz (2013) provided a general design scheme for minimum volume confidence regions under prior knowledge on the target parameter, which they applied to the problem of confidence intervals for a probability p. Prior knowledge on p was expressed by a beta distribution. We compare the scheme with the Bayesian HPD credibility intervals by imposing a beta prior and analyse the performance of the intervals in terms of coverage probability and length. We put focus on right-skewed beta densities with a high probability mass close to 0, which play an important role e. g. in the context of auditing.
  • Electrical Load Forecasting by Exponential Smoothing with Covariates

    Authors: Kristina Lurz (University of Würzburg), Rainer Göb (University of Würzburg), Antonio Pievatolo (CNR-IMATI)
    Primary area of focus / application: Other: Energy Markets
    Keywords: Exponential smoothing with covariates, Time series, Electrical load forecasting, Multiple seasonalities, Temperature effects
    Submitted at 29-May-2013 17:26 by Kristina Lurz
    Accepted
    16-Sep-2013 15:05 Electrical Load Forecasting by Exponential Smoothing with Covariates
    In the past, studies in short-term electrical load forecasting have been rather sceptical on the use of meteorological covariates like temperature for short-term forecasting purposes. The main reasons were time delays in data provision and the poor precision of meteorological forecasts. Both arguments have lost their impact, as new recent studies have shown. We explore the use of meteorological covariates in short-term load forecasting based on the rather new method of exponential smoothing with covariates (ESCov). The existing ESCov model is refined by including multiple
    seasonalities. The method is empirically explored in the hourly prediction of the electrical consumption of customers from provinces of an Italian region.
  • A Local Syndromic Surveillance Initiative Using Disease Count Data of a Pediatric Emergency Department

    Authors: Eralp Dogu (Department of Statistics Mugla Sitki Kocman University), Zeynep Filiz Eren-Dogu (Department of Computer Engineering Mugla Sitki Kocman University), Murat Anil (Department of Pediatrics, Izmir Tepecik Training and Research Hospital), Caner Alparslan (Department of Pediatrics, Izmir Tepecik Training and Research Hospital), Engin Kose (Department of Pediatrics, Izmir Tepecik Training and Research Hospital), Ender Can (Department of Pediatrics, Izmir Tepecik Training and Research Hospital)
    Primary area of focus / application: Process
    Keywords: Syndromic Surveillance, Disease Outbreak, Outbreak Detection Algorithms, Pediatric Emergency
    Submitted at 29-May-2013 20:41 by Eralp Dogu
    Accepted
    17-Sep-2013 16:35 A Local Syndromic Surveillance Initiative Using Disease Count Data of a Pediatric Emergency Department
    The main purpose of syndromic surveillance is early detection of syndromic disease outbreaks. A syndromic surveillance system mostly depends upon data provided by medical sources. Chief complaint and diagnosis data gathered from emergency departments may be effectively used in order to detect a disease outbreak. The aim of this research is to indicate how emergency data can help medical professionals get prepared for a disease outbreak. The study is a retrospective analysis of the pediatric emergency data of a single urban medical facility in Turkey from July 2011 through July 2012. We evaluate the data in several ways. First we use the daily counts of certain diseases. Using time series of these counts and applying surveillance algorithms we initiate a surveillance system for this medical facility. The needs, challenges and solutions to this implementation are discussed in this study.
  • Multivariate Statistical Process Control to Monitor Hypertensive Patients

    Authors: Eralp Dogu (Department of Statistics Mugla Sitki Kocman University), Min-Jung Kim (Department of Mechanical Engineering University of Maryland), Harriet Black Nembhard (Department of Industrial and Manufacturing Engineering The Pennsylvania State University)
    Primary area of focus / application: Process
    Keywords: multivariate control charts, statistical monitoring, chronic hypertension, blood pressure data
    Submitted at 29-May-2013 20:52 by Eralp Dogu
    Accepted
    17-Sep-2013 10:30 Multivariate Statistical Process Control to Monitor Hypertensive Patients
    Multivariate control charts (MVCC) are essential to the statistical monitoring of processes that produce interrelated quality metrics. In this study, we develop new algorithms for MVCC that enable us to analyze many factors and yet evaluate and interpret individual factors so that users can take mitigating action. We consider the implementation and performance of the algorithms in light of the challenges faced in monitoring patients with chronic hypertension. We support our discussion with a case study using blood pressure data from patients with chronic hypertension.