Skip to main content
SpringerLink
Log in

Decision Analytics mit Heatmap-Visualisierung von mehrschrittigen Ensembledaten

Eine Anwendung von Unsicherheitsmodellierung für Historical Consistent Neural Network und andere Prognosetechniken

Decision Analytics with Heatmap-Visualization for Multi-step Ensemble Data

An Application of Uncertainty Modeling to Historical Consistent Neural Network and Other Forecasts

  • Aufsatz
  • Published:
WIRTSCHAFTSINFORMATIK

Zusammenfassung

Heutige in verschiedenen Informationssystemen integrierte Prognosetechniken nutzen oft Ensembles zur Darstellung verschiedener zukünftiger Szenarien. Die Aggregation dieser Prognosen stellt eine anspruchsvolle Aufgabe da: Bei der Nutzung von Mittelwert und Median (gängige Praxis) gehen wichtige Informationen verloren, vor allem wenn die zugrunde liegende Verteilung zu jedem Schritt multimodal ist. Um dies zu vermeiden präsentieren wir einen Heatmap-Visualisierungsansatz. Visuell ist eine einfache Unterscheidung zwischen Bereichen mit hoher Aktivität (hohe Wahrscheinlichkeit der Realisierung) und solchen mit niedriger Aktivität möglich. Diese Form der Darstellung ermöglicht eine Identifikation von sich aufspaltenden Pfaden im Prognoseensemble und schafft dadurch eine „dritte Alternative“ im Entscheidungsraum. Die meisten Prognosesysteme bieten nur Ergebnisse „auf“ oder „ab“ an. Die vorgestellte Heatmap-Visualisierung führt zusätzlich ein Ergebnis „weiß nicht“ ein. Durch Blick auf die Heatmap können somit Bereiche identifiziert werden, in denen sich das zugrunde liegende Prognosemodell nicht sicher ist über den zukünftigen Ausgang. Wir präsentieren einen Softwareprototyp zur Unterstützung von Entscheidern durch eine interaktive Visualisierung und diskutieren den Informationsgewinn durch die Nutzung. Der Prototyp wurde bereits anderen Forschern und Praktikern präsentiert und mit diesen diskutiert.

Abstract

Today’s forecasting techniques, which are integrated into several information systems, often use ensembles that represent different scenarios. Aggregating these forecasts is a challenging task: when using the mean or median (common practice), important information is lost, especially if the underlying distribution at every step is multimodal. To avoid this, the authors present a heatmap visualization approach. It is easy to visually distinguish regions of high activity (high probability of realization) from regions of low activity. This form of visualization allows to identify splitting paths in the forecast ensemble and adds a “third alternative” to the decision space. Most forecast systems only offer “up” or “down”: the presented heatmap visualization additionally introduces “don’t know”. Looking at the heatmap, regions can be identified in which the underlying forecast model cannot predict the outcome. The authors present a software prototype with interactive visualization to support decision makers and discuss the information gained by its use. The prototype has already been presented to and discussed with researchers and practitioners.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Abb. 1
Abb. 2
Abb. 3
Abb. 4
Abb. 5
Abb. 6

Literatur

  • Abel GJ (2012) The fanplot package for R. http://gjabel.wordpress.com/2012/08/13/the-fanplot-package-for-r/. Abruf am 2013-10-21

  • Abel GJ (2013) Fanplot: visualisation of sequential probability distributions using fan charts. http://cran.r-project.org/web/packages/fanplot/index.html. Abruf am 2013-10-21

  • Aigner W, Bertone A, Miksch S, Tominski C, Schumann H (2007) Towards a conceptual framework for visual analytics of time and time-oriented data. In: Henderson SG, Biller B, Hsieh MH, Shortle J, Tew JD, Barton RR (Hrsg) Proc 2007 winter simulation conference, Washington, S 721–729

    Chapter  Google Scholar 

  • Andrienko G, Andrienko N (2005) Visual exploration of the spatial distribution of temporal behaviors. In: Proc ninth international conference on information visualisation. IEEE Computer Society, London, S 799–806

    Chapter  Google Scholar 

  • Andrienko G, Andrienko N, Mladenov M, Mock M, Poelitz C (2010) Extracting events from spatial time series. In: Proc 14th international conference on information visualisation. IEEE Computer Society, London, S 48–53

    Chapter  Google Scholar 

  • Bade R, Schlechtweg S, Miksch S (2004) Connecting time-oriented data and information to a coherent interactive visualization. In: Proc of ACM conference on human factors in computing systems (CHI’04). ACM, New York, S 105–112

    Google Scholar 

  • Buono P, Plaisant C, Simeone A, Aris A, Shneiderman B, Shmueli G, Jank W (2007) Similarity-based forecasting with simultaneous previews: a river plot interface for time series forecasting. In: Proc 11th international conference information visualization (2007). IEEE Computer Society, Zurich. doi:10.1109/IV.2007.101

    Google Scholar 

  • Elder R, Kaperanios G, Taylor T, Yates T (2005) Assessing the MPC’s fan charts. Bank of England quarterly bulletin: autumn 2005. Bank of England, London, S 326–348

    Google Scholar 

  • Feng D, Kwock L, Lee Y, Taylor RM II (2010) Matching visual saliency to confidence in plots of uncertain data. IEEE Transactions on Visualization and Computer Graphics 16(6):980–989

    Article  Google Scholar 

  • Hansen BE (2008) Least-squares forecast averaging. Journal of Econometrics 146(2):342–350. doi:10.1016/j.jeconom.2008.08.022

    Article  Google Scholar 

  • Hao MC, Janetzko H, Sharma RK, Dayal U, Keim DA, Castellanos M (2009) Poster: visual prediction of time series. In: Proc IEEE symposium on visual analytics science and technology (VAST 2009), S 229–230

    Chapter  Google Scholar 

  • Hevner AR, March ST, Park J, Ram S (2004) Design science in information systems research. MIS Quarterly 28(1):75–105

    Google Scholar 

  • May R, Hanrahan P, Keim DA, Shneiderman B, Card S (2010) The state of visual analytics: views on what visual analytics is and where it is going. In: Proc IEEE symposium on visual analytics science and technology (VAST), Salt Lake City, S 257–259

    Google Scholar 

  • Potter K, Wilson A, Bremer PT, Williams D, Doutriaux C, Pascucci V, Johnson CR (2009a) Ensemble-vis: a framework for the statistical visualization of ensemble data. In: Proc international conference on data mining workshops, S 233–240

    Google Scholar 

  • Potter K, Wilson A, Bremer PT, Williams D, Pascucci V, Johnson C (2009b) A flexible approach for the statistical visualization of ensemble data. In: Proc IEEE ICDM workshop on knowledge discovery from climate data, Miami

    Google Scholar 

  • Potter K, Rosen P, Johnson CR (2012) From quantification to visualization: a taxonomy of uncertainty visualization approaches. In: Dienstfrey AM, Boisvert RF (Hrsg) Uncertainty quantification in scientific computing. IFIP advances in information and communication technology, vol 377, S 226–249. doi:10.1007/978-3-642-32677-6_15

    Chapter  Google Scholar 

  • Raymer J, Abel GJ, Rogers A (2012) Does specification matter? Experiments with simple multiregional probabilistic population projections. Environment and Planning A 44(11):2664–2686. doi:10.1068/a4533

    Article  Google Scholar 

  • Savikhin A, Lam HC, Fisher B, Ebert DS (2011) An experimental study of financial portfolio selection with visual analytics for decision support. In: Proc 44th Hawaii international conference on system sciences (HICSS). doi:10.1109/HICSS.2011.54

    Google Scholar 

  • Thomas JJ, Cook KA (2006) A visual analytics agenda. Computer Graphics and Applications, IEEE 46(1):10–13

    Article  Google Scholar 

  • Uchida Y, Itoh T (2009) A visualization and level-of-detail control technique for large scale time series data. In: Proc 13th international conference information visualisation, Barcelona, S 80–85. doi:10.1109/IV.2009.33

    Google Scholar 

  • von Mettenheim HJ, Breitner MH (2010) Robust decision support systems with matrix forecasts and shared layer perceptrons for finance and other applications. In: Proc ICIS 2010, St. Louis, Paper 83

    Google Scholar 

  • von Mettenheim HJ, Köpp C, Breitner MH (2012) Visualizing forecasts of neural network ensembles. In: Klatte D, Lüthi HJ, Schmedders K (Hrsg) Operations research proceedings 2011, Zurich, S 573–578. doi:10.1007/978-3-642-29210-1_91

    Chapter  Google Scholar 

  • Welch I (2001) The equity premium consensus forecast revisited. Cowles Foundation discussion paper No 1325, University of California, Los Angeles (UCLA), National Bureau of Economic Research (NBER)

  • Yagi S, Uchida Y, Itoh T (2012) A polyline-based visualization technique for tagged time-varying data. In: Proc 16th international conference on information visualisation, Montpellier, S 106–111. doi:10.1109/IV.2012.28

    Google Scholar 

  • Zhang GP, Berardi VL (2001) Time series forecasting with neural network ensembles: an application for exchange rate prediction. The Journal of the Operational Research Society 52(6):652–664

    Article  Google Scholar 

  • Zimmermann HG, Grothmann R, Tietz C, von Jouanne-Diedrich H (2010) Market modeling, forecasting and risk analysis with historical consistent neural networks. In: Selected papers of the annual international conference of the German Operations Research Society, Munich, S 531–536

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Institut für Wirtschaftsinformatik, Leibniz Universität Hannover, Königsworther Platz 1, 30167, Hannover, Deutschland

    Cornelius Köpp,  Hans-Jörg von Mettenheim &  Michael H. Breitner

Authors
  1. Cornelius Köpp
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Hans-Jörg von Mettenheim
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Michael H. Breitner
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Hans-Jörg von Mettenheim.

Additional information

Angenommen nach zwei Überarbeitungen durch die Herausgeber des Schwerpunktthemas.

This article is also available in English via http://www.springerlink.com and http://www.bise-journal.org: Köpp C, von Mettenheim H-J, Breitner MH (2014) Decision Analytics with Heatmap Visualization for Multi-step Ensemble Data. An Application of Uncertainty Modeling to Historical Consistent Neural Network and Other Forecasts. Bus Inf Syst Eng. doi: 10.1007/s12599-014-0326-4.

Zusätzliche Information

Ergänzendes Onlinematerial kann unter folgendem Link abgerufen werden.

(PDF 2.9 MB)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Köpp, C., von Mettenheim, HJ. & Breitner, M.H. Decision Analytics mit Heatmap-Visualisierung von mehrschrittigen Ensembledaten. Wirtschaftsinf 56, 147–157 (2014). https://doi.org/10.1007/s11576-014-0417-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11576-014-0417-3

Schlüsselwörter

Keywords

Search

Navigation