A neural network based dynamic forecasting model for Trend Impact Analysis
| Autor: | Nedaa Agami, Hisham El-Shishiny, Amir F. Atiya, Mohamed Saleh |
|---|---|
| Rok vydání: | 2009 |
| Předmět: |
Artificial neural network
Mechanism (biology) business.industry Process (engineering) Computer science Machine learning computer.software_genre Futures studies Management of Technology and Innovation Probabilistic forecasting Artificial intelligence State (computer science) Business and International Management business computer Applied Psychology |
| Zdroj: | Technological Forecasting and Social Change. 76:952-962 |
| ISSN: | 0040-1625 |
| DOI: | 10.1016/j.techfore.2008.12.004 |
| Popis: | Trend Impact Analysis is a simple forecasting approach, yet powerful, within the Futures Studies paradigm. It utilizes experts' judgements to explicitly deal with unprecedented future events with varying degrees of severity in generating different possibilities (scenarios) of how the future might unfold. This is achieved by modifying a surprise-free forecast according to events' occurrences based on a Monte-Carlo simulation process. Yet, the current forecasting mechanism of TIA is static. This paper introduces a new approach for constructing TIA by using a dynamic forecasting model based on neural networks. This new approach is designed to enhance the TIA prediction process. It is expected that such a dynamic mechanism will produce more robust and reliable forecasts. Its idea is novel, beyond state of the art and its implementation is the main contribution of this paper. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect