Details
| Original language | English |
|---|---|
| Pages (from-to) | 131-140 |
| Number of pages | 10 |
| Journal | Bus. Inf. Syst. Eng. |
| Volume | 6 |
| Issue number | 3 |
| Publication status | Published - Jun 2014 |
Abstract
With today's computing power, it is easy to generate huge amounts of data. The real challenge lies in adequately condensing the data in decision making processes. Here, the focus is on ensemble data that typically arises when distributions of forecasts are generated for several time steps in the future. Often a distribution is aggregated by taking an ensemble's mean or median. This results in a single line that is easy to interpret. However, this single line may be seriously misleading when the ensemble splits into two or more different bundles. The mean or median may also lie in a region where there are only very few ensemble members. To remedy this, a heatmap visualization to better represent ensemble data for decision analytics is proposed. Heatmap visualization provides an intuitive way to identify regions of high and low activity. The regions are color-coded according to the (weighted) number of ensemble members in a specific region.
ASJC Scopus subject areas
- Computer Science(all)
- Information Systems
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Bus. Inf. Syst. Eng., Vol. 6, No. 3, 06.2014, p. 131-140.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - Decision analytics with heatmap visualization for multi-step ensemble data
AU - Köpp, Cornelius
AU - Mettenheim, Hans-Jörg von
AU - Breitner, Michael H.
N1 - Copyright: Copyright 2014 Elsevier B.V., All rights reserved.
PY - 2014/6
Y1 - 2014/6
N2 - With today's computing power, it is easy to generate huge amounts of data. The real challenge lies in adequately condensing the data in decision making processes. Here, the focus is on ensemble data that typically arises when distributions of forecasts are generated for several time steps in the future. Often a distribution is aggregated by taking an ensemble's mean or median. This results in a single line that is easy to interpret. However, this single line may be seriously misleading when the ensemble splits into two or more different bundles. The mean or median may also lie in a region where there are only very few ensemble members. To remedy this, a heatmap visualization to better represent ensemble data for decision analytics is proposed. Heatmap visualization provides an intuitive way to identify regions of high and low activity. The regions are color-coded according to the (weighted) number of ensemble members in a specific region.
AB - With today's computing power, it is easy to generate huge amounts of data. The real challenge lies in adequately condensing the data in decision making processes. Here, the focus is on ensemble data that typically arises when distributions of forecasts are generated for several time steps in the future. Often a distribution is aggregated by taking an ensemble's mean or median. This results in a single line that is easy to interpret. However, this single line may be seriously misleading when the ensemble splits into two or more different bundles. The mean or median may also lie in a region where there are only very few ensemble members. To remedy this, a heatmap visualization to better represent ensemble data for decision analytics is proposed. Heatmap visualization provides an intuitive way to identify regions of high and low activity. The regions are color-coded according to the (weighted) number of ensemble members in a specific region.
UR - http://www.scopus.com/inward/record.url?scp=84905818005&partnerID=8YFLogxK
U2 - 10.1007/s12599-014-0326-4
DO - 10.1007/s12599-014-0326-4
M3 - Article
VL - 6
SP - 131
EP - 140
JO - Bus. Inf. Syst. Eng.
JF - Bus. Inf. Syst. Eng.
IS - 3
ER -