Listing 1 - 4 of 4 |
Sort by
|
Choose an application
The desire to have access to information at all times, while at the same time being informed automatically is the driving force behind this thesis. The goal is to bring an existing dashboard to mobile devices and to research algorithms that find outliers in timeseries.Because the dasboard is only used by a small number of people it is important to find a cheap and maintainable solution for the mobile dashboard. This has been achieved by using the HTML5 standard in combination with SVG and media queries to transform the dashboard into a webapp.When outliers are detected in specific metrics the qualified persons are alerted. However, the current solution generates too much alerts. These metrics are made of hourly measurements of activities on the products of Massive Media. With linear regression the dataset is decomposed in a trend and seasonal component. The remainder is assumed to fit a normal distribution. Outliers can then be found and reported by using a robust measurement for variation like the Median Absolute Deviation (MAD).This method finds outliers in the data that are undetectable with the naked eye and finds 40% less outliers than the currently used method. Vandaag de dag wordt er verwacht dat informatie op elk moment beschikbaar is, waarbij men tegelijkertijd proactief wordt geïnformeerd. Het doel van deze scriptie is dan ook een bestaand dashboard op mobiele toestellen brengen en algoritmen onderzoeken die uitschieters in data ontdekken.Door het beperkt publiek dat het dashboard gebruikt is het van belang een goedkope en onderhoudbare oplossing te vinden. Dit wordt bereikt door gebruik te maken van de nieuwe HTML5 standaard in combinatie met Scalable Vector Graphics (SVG) en media queries om het dashboard om te vormen tot een webapp.Wanneer uitschieters in data voorkomen worden de bevoegde personen verwittigd. De huidige oplossing genereert echter te veel alerts. De datasets bestaan uit uurgebonden metingen van activiteiten op de producten van Massive Media. Door deze te ontbinden in een trend- en seizoensbeweging met behulp van lineaire regressie, wordt de restcomponent bepaald waarvan we veronderstellen dat die normaal verdeeld is. Uitschieters kunnen vervolgens met behulp van robuste statistische methoden, zoals de Mediaan Absolute Deviatie (MAD), ontdekt en gerapporteerd worden.De nieuwe implementatie vindt uitschieters in de data die onmogelijk te vinden zijn met het blote oog en geeft ongeveer 40% minder uitschieters dan de bestaande methode.
Algoritme - algorithm. --- CSS - CSS. --- Dashboard. --- HTML5. --- JavaScript - JavaScript. --- Lineaire regressie. --- MAD. --- Media queries. --- Optimalisatie - optimization. --- Python. --- SVG. --- Webapplicatie - web application.
Choose an application
Econometrics --- 330.015195 --- 303.5 --- AA / International- internationaal --- Economics, Mathematical --- Statistics --- Theorie van correlatie en regressie. (OLS, adjusted LS, weighted LS, restricted LS, GLS, SLS, LIML, FIML, maximum likelihood). Parametric and non-parametric methods and theory (wiskundige statistiek)
Choose an application
1. Gegevens - 2. Kans en inferentie - 3. Onderwerpen binnen inferentie - 4. Supplement
519.2 --- statistiek (wiskunde) --- 31 <035> --- 519.25 --- 311 --- #KVHB:Statistiek --- statistiek --- 311 Statistische methoden --- Statistische methoden --- 519.25 Statistical data handling --- Statistical data handling --- 31 <035> Statistieken --(algemeen z.o.{519.2})--Grote handboeken. Compendia --- Statistieken --(algemeen z.o.{519.2})--Grote handboeken. Compendia --- Waarschijnlijkheidsrekening. Mathematische of wiskundige statistiek --- Mathematical statistics --- Statistiques Statistiek --- statistieken --- Kansrekening --- Statistiek --- Regressie --- kansrekening --- variantieanalyse --- regressie --- Variantieanalyse --- 301.2 --- sociale en economische statistiek, theorie en methodiek --- Mathematical statistics. --- Wiskundige statistiek --- PXL-Central Office 2018 --- Documenten : Statistische documenten --- Documents : Documents statistiques --- Statistiques Statistiek.
Choose an application
This is a practical introduction to multilevel analysis suitable for all those doing research. Most books on multilevel analysis are written by statisticians, and they focus on the mathematical background. These books are difficult for non-mathematical researchers. In contrast, this volume provides an accessible account on the application of multilevel analysis in research. It addresses the practical issues that confront those undertaking research and wanting to find the correct answers to research questions. This book is written for non-mathematical researchers and it explains when and how to use multilevel analysis. Many worked examples, with computer output, are given to illustrate and explain this subject. Datasets of the examples are available on the internet, so the reader can reanalyse the data. This approach will help to bridge the conceptual and communication gap that exists between those undertaking research and statisticians.
Biomathematics. Biometry. Biostatistics --- Mathematical statistics --- Model, Statistical --- Models, Binomial --- Models, Polynomial --- Statistical Model --- Probabilistic Models --- Statistical Models --- Two-Parameter Models --- Binomial Model --- Binomial Models --- Model, Binomial --- Model, Polynomial --- Model, Probabilistic --- Model, Two-Parameter --- Models, Probabilistic --- Models, Two-Parameter --- Polynomial Model --- Polynomial Models --- Probabilistic Model --- Two Parameter Models --- Two-Parameter Model --- Analysis of variance --- Medical statistics --- Medicine --- Biomedical Research --- Models, Statistical --- Multivariate Analysis --- Statistics as Topic --- multivariaat --- regressie-analyse --- wiskundige statistiek --- Analyses, Multivariate --- Analysis, Multivariate --- Multivariate Analyses --- Biomedical research --- Medical research --- Health --- Health statistics --- Statistics --- ANOVA (Analysis of variance) --- Variance analysis --- Experimental design --- Research --- methods --- Statistical methods --- Models, Statistical. --- Multivariate Analysis. --- Analysis of variance. --- Medical statistics. --- methods. --- Research. --- Health Workforce --- Canonical Correlation --- Canonical Correlations --- Correlation, Canonical --- Health Sciences --- General and Others
Listing 1 - 4 of 4 |
Sort by
|