SCM에서 SARIMA와 LSTM의 앙상블 학습

Alternative Title
Hanyul Ryu
Author(s)
류한열
Alternative Author(s)
Hanyul Ryu
Advisor
권순선
Department
일반대학원 데이터사이언스학과
Publisher
The Graduate School, Ajou University
Publication Year
2019-08
Language
kor
Abstract
공급사슬관리(Supply Chain Management; SCM)이란공급망전체를하나의통합된개체로보고이를최적화하고자하는경영방식이다.운영의효율이라는관점에서는빅데이터를활용하여시장의수요예측,실시간경로최적화,창고배치최적화등의프로세스효율을증대하고운영비용을감소시킬수있다.하지만계절성,물품에대한유행,회사 사업마케팅전략,이벤트등의외부사회현상으로인하여시장의예측은매우힘든일이다.본논문에서는SCM의계절성이반영된데이터의수요예측을 위해서데이터의전처리방법과특정단위기간의예측을 위한분류자(Classifier)의설정을설명하고계절성자기회귀누적이동평균(Seasonal Autoregressive IntergratedMoving Average; SARIMA)모델과장단기 기억(Long-Short Term Memory;LSTM)모델간의앙상블학습을통한예측방법을설명하여수요예측모델을구축하고자한다.
URI
https://dspace.ajou.ac.kr/handle/2018.oak/15438
Fulltext

Appears in Collections:
Graduate School of Ajou University > Department of Data Science > 3. Theses(Master)
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse