• Industrial Engineering and Management Systems
• /
• 제14권2호
• /
• pp.196-209
• /
• 2015

Retail sales forecast is a special area of forecasting. Its unique characteristics call for unique data models and treatment, and unique forecasting processes. In this paper, we will address lessons learned and challenges encountered in retail sales forecast from a practical and technical perspective. In particular, starting with the data models of retail sales data, we proceed to address issues existing in estimating and processing each component in the data model. We will discuss how to estimate the multi-seasonal cycles in retail sales data, and the limitations of the existing methodologies. In addition, we will talk about the distinction between business events and forecast events, the methodologies used in event detection and event effect estimation, and the difficulties in compound event detection and effect estimation. For each of the issues and challenges, we will present our solution strategy. Some of the solution strategies can be generalized and could be helpful in solving similar forecast problems in different areas.

• 산업경영시스템학회지
• /
• 제17권31호
• /
• pp.1-10
• /
• 1994

The purpose of this study is to search for the most effective forecasting model for condenser with independent demand among the quantitative methods such as Brown's exponential smoothing method, Box-Jenkins method, and multiple regression analysis method. The criterion for the comparison of the above models is mean squared error(MSE). The fitting results of these three methods are as follows. 1) Brown's exponential smoothing method is the simplest one, which means the method is easy to understand compared to others. But the precision is inferior to other ones. 2) Box-Jenkins method requires much historic data and takes time to get to the final model, although the precision is superior to that of Brown's exponential smoothing method. 3) Regression method explains the correlation between parts with similiar demand pattern, and the precision is the best out of three methods. Therefore, it is suggested that the multiple regression method is fairly good in precision for forecasting our item and that the method is easily applicable to practice.

• 부동산연구
• /
• 제27권3호
• /
• pp.41-50
• /
• 2017

변동성 모형을 이용한 국내의 주택가격에 대한 기존의 연구에서는 변동성모형을 어떻게 주택시장분석에 적용할 수 있는지를 보여주고 있지만 최근 국내의 지역주택시장들에 나타나는 유의미한 변화를 반영하는데는 한계가 존재할 수 밖에 없다. 본 연구에서는 변동성모형을 적용하여 전국의 각 지역별 주택시장을 분석하고 이를 통해 미래의 지역별 주택시장의 가격변동을 실제적으로 예측하였다. AR(1)-ARCH(1), AR(1)-GARCH(1,1), AR(1)-EGARCH(1,1,1) 모형을 통하여 지역주택시장에 ARCH 및 GARCH효과가 존재하는 것을 확인하였다. 그리고 각 지역의 예측력을 비교하여 지역별 최적예측모형을 선정하였으며, 이러한 지역별 최적모형의 선정이 실제적으로 어떻게 이용될 수 있는지를 보여주기 위하여 2017년 하반기의 각 지역주택시장의 가격변동을 선정된 지역별 최적모형을 이용하여 예측하였다.

• Journal of Preventive Medicine and Public Health
• /
• 제21권1호
• /
• pp.195-206
• /
• 1988

Two major issues of the blood bank management are quality assurance and inventory control. Recently, in Korea blood donation has gained popularity increasingly to allow considerable improvement of the quality assurance with respect to blood collection, transportation, storage, component preparation skills and hematological tests. Nevertheless the inventory control, the other issue of blood bank management, has been neglected so far. For the supply of blood by donation barely meets the demand, the blood bank policy on the inventory control has been 'the more the better.' The shortage itself by no means unnecessitate inventory control. In fact, in spite of shortage, no small amount of blood is outdated. The efficient blood inventory control makes it possible to economize the blood usage in the practice of state-of-the-art medical care. For the efficient blood inventory control in Korean hospitals, this tudy is to develop formulae forecasting the standard blood inventory level and suggest a set of policies improving the blood inventory control. For this study informations of $A^+$ whole bloods and packed cells inventory control were collected from a University Hospital and the Central Blood Bank of the Korean Red Cross. Using this informations, 1,461 daily blood inventory records were formulated.48 varieties of blood inventory control environment were identified on the basis of selected combinations of 4 inventory control variables-crossmatch, transfusion, inhospital donation and age of bloods from external supply. In order to decide the optimal blood inventory level for each environment, simulation models were designed to calculate the measures of performance of each environment. After the decision of 48 optimal blood inventory levels, stepwise multiple regression analysis was started where the independent variables were 4 inventory control variables and the dependent variable was optimal inventory level of each environment. Finally the standard blood inventory level decision rule was developed using the backward elimination procedure to select the best regression equation. And the effective alternatives of the issuing policy and crossmatch release period were suggested according to the measures of performance under the condition of the standard blood inventory level. The results of this study' were as follows ; 1. The formulae to calculate the standard blood inventory level($S^*$)was $S^*=2.8617X(d)^{0.9342}$ where d is the mean daily crossmatch(demand) for a blood type. 2. The measures of performace - outdate rate, average period of storage, mean age of transfused bloods, and mean daily available inventory level - were improved after maintenance of the standard inventory level in comparison with the present system. 3. Issuing policy of First In-First Out(FIFO) decreased the outdate rate, while Last In-First Out(LIFO) decreased the mean age of transfused bloods. The decrease of the crossmatch release period reduced the outdate rate and the mean age of transfused bloods.

• 지식경영연구
• /
• 제23권1호
• /
• pp.89-109
• /
• 2022

최근 COVID-19, 동학개미운동 등 투자환경의 변화로 시스템 처리 허용 수준을 상회하는 트랜잭션이 발생하고 이로 인해 전산장애가 자본시장에서 빈번하게 나타나고 있다. 자본시장 IT시스템들은 장애 영향도가 매우 큰 시스템들로서, 2020년에 예측하지 못한 큰 규모의 트랜잭션이 상당한 기간 유입되어 전산장애가 급증하였다. 다수의 기업들이 높은 수준의 IT시스템 용량계획 정책을 유지하고 있던 상황임에도 불구하고, 이를 상회하는 트랜잭션이 유입된 것은 용량계획에 대한 새로운 접근 방법이 필요함을 시사하고 있다. 이에 본 연구는 다양한 머신러닝 기법을 활용하여 자본시장 IT시스템 용량계획 모델들을 개발하고 성능을 비교 분석한다. 또한, 동학개미운동과 같이 예측하기 힘든 투자자의 행동을 반영할 수 있는 심리지수를 예측에 활용함으로써 용량계획 모델의 성능을 높인다. COVID-19 기간을 포함한 실증데이터를 이용하여 본 연구에서 개발한 용량계획 모델은 실무에서 활용 가능한 수준의 높은 성능과 안정성을 가질 수 있다. 본 연구는 기업의 비용 효율성과 IT시스템 용량 변경에 수반되는 운영상의 제약을 모두 고려한 최적의 파라미터를 제시하였는데, 이것은 자본시장 도메인에서 유용하게 사용될 수 있다. 또한, 본 연구는 투자자의 심리를 반영하는 심리지수가 IT 시스템 용량계획에 중요한 예측요인이 될 수 있는 것을 입증함으로써, 심리지수가 다양한 수요예측에 적극적으로 활용될 수 있음을 보여준다.