DEMAND ESTIMATING METHOD AND INVENTORY MANAGEMENT SYSTEM USING MACHINE LEARNING

• Main Authors: PARK, DO GUN, JUNG, JONG PIL, KIM, MYUNG SOO
• Format: Patent
• Language: English; Korean
• Published: 18.06.2021
• Subjects: CALCULATING; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

Provided, in the present invention, is a demand forecasting method using machine learning comprising: a step of preprocessing by acquiring manufacturing management data which comprises past company's shipment data; a step of deriving a forecast value after extracting only shipment and time data from the preprocessed data, and extracting highly correlated data from the remaining data; a step of merging the derived and extracted data and dividing the merged data into a plurality of batches of the same size; a step of learning a plurality of disposed datasets; and a step of estimating demand and appropriate inventory based on the learning result. Therefore, the present invention is capable of reducing a problem of reliability for which is a chronic problem for forecasting demand. 본 발명은 과거 기업의 출고데이터를 포함한 생산관리 데이터를 획득하여 전처리하는 단계, 전처리된 데이터에서 출고, 시간 데이터만 추출한 후 예측값을 도출하고, 남은 데이터에서 상관관계가 높은 데이터를 추출하는 단계, 상기 도출 및 추출된 데이터를 병합하고 상기 병합된 데이터를 동일한 크기의 복수개의 배치에 나누는 단계, 복수개의 배치된 데이터셋을 학습하는 단계, 상기 학습 결과에 기반하여 수요 및 적정재고를 예상하는 단계를 포함하는 기계적 학습을 이용한 수요예측 방법을 제공한다.