제약적 NLS 방법을 이용한 출시 초기 신제품의 중장기 수요 예측 방안
Constrained NLS Method for Long-term Forecasting with Short-term Demand Data of a New Product
홍정식(서울산업대 산업정보시스템공학과); 구훈영(충남대학교)
38권 1호, 45~59쪽
초록
A long-term forecasting method for a new product in early stage of diffusion is proposed. The method includes a constrained non-linear least square estimation with the logistic diffusion model. The constraints would be critical market informations such as market potential, peak point, and take-off. Findings on 20 cases having almost full life cycle are that (i) combining any market information improves the forecasting accuracy, (ii) market potential is the most stable information, and (iii) peak point and take-off information have negative effect in case of overestimation.
Abstract
A long-term forecasting method for a new product in early stage of diffusion is proposed. The method includes a constrained non-linear least square estimation with the logistic diffusion model. The constraints would be critical market informations such as market potential, peak point, and take-off. Findings on 20 cases having almost full life cycle are that (i) combining any market information improves the forecasting accuracy, (ii) market potential is the most stable information, and (iii) peak point and take-off information have negative effect in case of overestimation.
- 발행기관:
- 한국경영과학회
- 분류:
- 경영학