Details
| Originalsprache | Englisch |
|---|---|
| Seiten | 478-489 |
| Seitenumfang | 12 |
| Publikationsstatus | Veröffentlicht - 2007 |
| Veranstaltung | 15th European Conference on Information Systems, ECIS 2007 - St. Gallen, Schweiz Dauer: 7 Juni 2007 → 9 Juni 2007 |
Konferenz
| Konferenz | 15th European Conference on Information Systems, ECIS 2007 |
|---|---|
| Land/Gebiet | Schweiz |
| Ort | St. Gallen |
| Zeitraum | 7 Juni 2007 → 9 Juni 2007 |
Abstract
Today self-organization and automatic usage of Artificial Neural Networks (ANN) are common in various applications for financial services providers. We analyze typical advantages and disadvantages of ANN and discuss the question: For which tasks ANN applications are most promising? We show that Intelligent Decision Support Systems (IDSS) based on ANN and Neurosimulators can support today's complex decision processes, e. g., investments or operation of a customer contact/call center. The focus is on supervised learning, here: ANN are trained with patterns from well-understood decision processes in the past. Then these ANN can benchmark a posteriori, forecast a priori or transfer knowledge to similar or analogous decision processes. Often efficient supervised learning needs advanced optimization algorithms, thin client solutions and low budget high performance computing, i. e. grid computing. Computations are realized with the neurosimulator FAUN (Fast Approximation with Universal Neural Networks), which is developed by the authors since the mid 1990's. We shortly present a long-term ANN interest rate forecasting model first. Then an ANN option/warrant market-pricing model and an ANN human-resource allocation model for contact/call centers are outlined briefly.
ASJC Scopus Sachgebiete
- Informatik (insg.)
- Information systems
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
2007. 478-489 Beitrag in 15th European Conference on Information Systems, ECIS 2007, St. Gallen, Schweiz.
Publikation: Konferenzbeitrag › Paper › Forschung › Peer-Review
}
TY - CONF
T1 - Intelligent decision support systems and neurosimulators
T2 - 15th European Conference on Information Systems, ECIS 2007
AU - Breitner, Michael H.
AU - Frank, Ller
AU - Simon, Nig
AU - Von Mettenheim, Hans Jörg
N1 - Copyright: Copyright 2012 Elsevier B.V., All rights reserved.
PY - 2007
Y1 - 2007
N2 - Today self-organization and automatic usage of Artificial Neural Networks (ANN) are common in various applications for financial services providers. We analyze typical advantages and disadvantages of ANN and discuss the question: For which tasks ANN applications are most promising? We show that Intelligent Decision Support Systems (IDSS) based on ANN and Neurosimulators can support today's complex decision processes, e. g., investments or operation of a customer contact/call center. The focus is on supervised learning, here: ANN are trained with patterns from well-understood decision processes in the past. Then these ANN can benchmark a posteriori, forecast a priori or transfer knowledge to similar or analogous decision processes. Often efficient supervised learning needs advanced optimization algorithms, thin client solutions and low budget high performance computing, i. e. grid computing. Computations are realized with the neurosimulator FAUN (Fast Approximation with Universal Neural Networks), which is developed by the authors since the mid 1990's. We shortly present a long-term ANN interest rate forecasting model first. Then an ANN option/warrant market-pricing model and an ANN human-resource allocation model for contact/call centers are outlined briefly.
AB - Today self-organization and automatic usage of Artificial Neural Networks (ANN) are common in various applications for financial services providers. We analyze typical advantages and disadvantages of ANN and discuss the question: For which tasks ANN applications are most promising? We show that Intelligent Decision Support Systems (IDSS) based on ANN and Neurosimulators can support today's complex decision processes, e. g., investments or operation of a customer contact/call center. The focus is on supervised learning, here: ANN are trained with patterns from well-understood decision processes in the past. Then these ANN can benchmark a posteriori, forecast a priori or transfer knowledge to similar or analogous decision processes. Often efficient supervised learning needs advanced optimization algorithms, thin client solutions and low budget high performance computing, i. e. grid computing. Computations are realized with the neurosimulator FAUN (Fast Approximation with Universal Neural Networks), which is developed by the authors since the mid 1990's. We shortly present a long-term ANN interest rate forecasting model first. Then an ANN option/warrant market-pricing model and an ANN human-resource allocation model for contact/call centers are outlined briefly.
KW - Artificial Neural Networks
KW - Contact/call centers
KW - Forecasting model
KW - Intelligent Decision Support Systems
KW - Neurosimulator
KW - Option pricing
UR - http://www.scopus.com/inward/record.url?scp=84869383602&partnerID=8YFLogxK
M3 - Paper
AN - SCOPUS:84869383602
SP - 478
EP - 489
Y2 - 7 June 2007 through 9 June 2007
ER -