Time series analysis and their optimization

2019 Fiscal Year Final Research Report

• Project/Area Number: 26870738
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Commerce; Management
• Research Institution: Setsunan University
• Principal Investigator: Higuchi Yuki 摂南大学, 経営学部, 准教授 (60552065)
• Project Period (FY): 2014-04-01 – 2020-03-31
• Keywords: 時系列解析 / 予測 / ブランド / Genetic Algorithm / 棚配置

Outline of Final Research Achievements

The following three themes of study related to time series analysis were executed. (1)Optimization in allocating goods to shop shelves utilizing Genetic Algorithm (2)Analysis of the matrix structure in brand selection (3)Improvement of the forecasting accuracy. The following results were obtained. (1)Exploring the problem that does not allow goods to be allocated in multiple shelves. Executing the applications to the sales data of cup noodles in the convenience store as a model. (2)Introduction of the rank classification based upon the consumers' perceived quality by utilizing the correspondence analysis. Executing a questionnaire investigation in the airline ticket purchasing case and confirming the upper shifts by the new method described above. (3)“The new index utilizing Mahalanobis’ generalized distance” and “Shortening the searching time of the optimal order of ARIMA model by utilizing Genetic Algorithm” are introduced and various forecasting for time series were conducted.

Free Research Field

経営情報、マーケティング

Academic Significance and Societal Importance of the Research Achievements

(1)本研究は店舗における商品配置問題を定式化し、その効果を数値として客観的に把握することを目的とする。店舗での陳列方法は売り上げに大きく影響するため、本手法の確立と精度の検証は商品販売戦略の展開に大きく資する。(2)本研究は消費者のブランド選択時における推移データを行列構造にて解明するものであり、法則化によりブランドの新製品投入タイミング、ポジショニングの判断などマーケティング戦略立案・検証に極めて有効に活用することができる。(3)在庫管理や見込み生産など、需要予測を必要とする分野・企業は非常にたくさん存在する。販売予測等、予測精度を上げることが切望される分野は多く、各方面への貢献度は高い。