Application of Chaos Theory to Sales Early Warning
碩士 === 國立中正大學 === 資訊管理所 === 94 === The importance of sales forecasting has grown magnificently in recent years. A precise forecasting not only helps the organizations to reduce costs of inventories but to avoid overstocking of capitals. The forecasting rule is typically constructed based on the know...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2006
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/01228291535149113122 |
| Summary: | 碩士 === 國立中正大學 === 資訊管理所 === 94 === The importance of sales forecasting has grown magnificently in recent years. A precise forecasting not only helps the organizations to reduce costs of inventories but to avoid overstocking of capitals. The forecasting rule is typically constructed based on the knowledge of historical records. When important features were neglected, the prediction model may derive results that are far from the truth. Hence it becomes a challenge for the managers to determine the most appropriate forecasting method especially for prediction of the future trends.
The chaos theory has been applied in many disciplines and has successfully resolved many difficult problems. In this thesis, we apply the chaos theory to sales forecasting by establishing a early warning mechanism. This mechanism aims to reduce the cost of inventories and to lower the pressure of overstocking of capitals. This early warning system also helps the organizations to take necessary precaution against upcoming problems.
We analyzed the sales activities of three food companies. The results showed that the products’ sale decreased when the lyapunov exponents went down in different levels 4 or 5 months ago. Our approach had 70% accuracy or higher in prediction. It showed better accuracy than that of moving average model. We also collected some relevant news to our research for organizations’ further references.
|
|---|