(An) intelligent decision support system for forecasting time series data : expert systems approach시계열예측을 위한 의사결정지원시스템:전문가시스템 접근법

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 395
  • Download : 0
Most existing types of forecasting software are limited in the sense that they require extensive knowledge about applied areas and statistical analysis. Furthermore they tend to ignore human factors in forecasting such as experience, craftsmanship, and subjective information. This thesis employs an AI (Artificial Intelligence) approach, in particular expect systems approach, which can solve forecasting problems more effectively and efficiently. We focus on the role of DSS (Decision Support System) in the first stage in forecasting, the model building process, and the third stage in forecasting, the forecast adjustment process. The domain of this study is forecasting of demand for oil products, which is represented in time series data. This study suggests a mixed approach to forecasting model building, which combines two extreme approaches, the knowledge-oriented approach and the data-oriented approach. The proposed approach accomplishes the integration of knowledge and data by direct integration or by indirect integration. The following issues are investigated: knowledge representation of modeling knowledge; inference mechanism for the modeling process; uncertainty management in the modeling process(represented in fuzziness). Time series models have served as a highly useful forecasting method, but are deficient in that they merely extrapolate from past patterns in data without taking into account expected future events and other qualitative factors. To overcome this limitation, forecasting experts in practice judgmentally adjust statistical forecasts. In order to incorporate the role of the forecasting expert``s judgment, we have developed a methodology which incorporates learning from historical judgmental adjustments through generalization and analogy, reasoning based on similar cases, and composing and decomposing the impacts of simultaneous judgmental events non-monotonically. Here, this is applied to the demand forecasting of oil products, for which five ...
Advisors
Lee, Jae-Kyuresearcher이재규researcher
Description
한국과학기술원 : 경영과학과,
Publisher
한국과학기술원
Issue Date
1990
Identifier
61565/325007 / 000835233
Language
eng
Description

학위논문(박사) - 한국과학기술원 : 경영과학과, 1990.2, [ vii, 96 p. ]

URI
http://hdl.handle.net/10203/43669
Link
http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=61565&flag=dissertation
Appears in Collection
MG-Theses_Ph.D.(박사논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0