資訊管理學報
范有寧;黃聖祐;陳靜枝;
頁: 51-75
日期: 2010/12
摘要: 在商業應用中，商品分類幾乎是所有使用與管理商品相關資訊活動的核心。企業普遍會為商品分類，以期透過此種分析模式與歸納方法可以有效提高商品的銷售量並增加企業營收。需求預測在供應鏈管理中扮演重要角色，良好的預測模式將幫助企業有效的存貨管理。然而，以往管理者以質性觀點所建立的商品分類架構無法完全適用於需求預測，貿然將銷售發展趨勢差異太大的商品歸為同一類，會導致類別商品的發展趨勢扭曲或模糊。因此，本研究將提出一以量化觀點為依據的自動化資料探勘模型，建立一套適合需求預測使用之商品分類架構，以達成提高商品銷售量與預測準確度的目標。 本研究以台灣地區知名的茶飲料商與知名連鎖藥妝店的銷售資料為實際案例，進行研究方法之驗證。經實驗結果發現，商品若擁有長期的銷售歷史，且具有明顯的長期趨勢與季節性波動，經由本研究所提出之分析方法可以有效判別商品之間的異同，並加以群集，以提升預測準確度。另外，本研究所提出之研究方法不受產業與商品限制，具有因應不同使用者之背景與未來應用產業的彈性。
關鍵字: 資料探勘;需求預測;商品分類;基因演算法;供應鏈管理;
Abstract: Product classification is the core of every information activity related to product management. Companies classify their products according to some attributes for different management functions. Within these functions, demand forecasting is the most critical oneof businessmanagement. Forecasts are essential to the business's decision making and planning processes. Better forecasting can contribute to better price structuring and better inventory management. However, it is a challenging problem owing to the volatility of demand which depends on many factors. Therefore, the study aims to design a classification scheme based on the quantitative characteristics of products and make it more suitable for demand forecasting. This study proposes a heuristic algorithm, called Data-Mining Aided Product Classification (DMAPC), to deal with aforementioned issues. DMAPC analyzes the sales records by using time-series analysis and searches the optimal product grouping result using GA-based heuristic algorithm. Accordingly, a new classification scheme is constructed by aforementioned processes. The proposed approach is applied to solve two real-world demand forecasting problems from a well-known cosmetic chain retailer and a prestigious tea retailer in Taiwan. The experimental results demonstrate that the proposed approach is proved to enhance the prediction accuracy effectively by applying DMAPC.
Keywords: Data Mining;Demand Forecasting;Product Classification;Genetic Algorithm;Supply Chain Management;
瀏覽次數: 16808 下載次數: 3782
引用 導入Endnote
頁: 51-75
日期: 2010/12
摘要: 在商業應用中，商品分類幾乎是所有使用與管理商品相關資訊活動的核心。企業普遍會為商品分類，以期透過此種分析模式與歸納方法可以有效提高商品的銷售量並增加企業營收。需求預測在供應鏈管理中扮演重要角色，良好的預測模式將幫助企業有效的存貨管理。然而，以往管理者以質性觀點所建立的商品分類架構無法完全適用於需求預測，貿然將銷售發展趨勢差異太大的商品歸為同一類，會導致類別商品的發展趨勢扭曲或模糊。因此，本研究將提出一以量化觀點為依據的自動化資料探勘模型，建立一套適合需求預測使用之商品分類架構，以達成提高商品銷售量與預測準確度的目標。 本研究以台灣地區知名的茶飲料商與知名連鎖藥妝店的銷售資料為實際案例，進行研究方法之驗證。經實驗結果發現，商品若擁有長期的銷售歷史，且具有明顯的長期趨勢與季節性波動，經由本研究所提出之分析方法可以有效判別商品之間的異同，並加以群集，以提升預測準確度。另外，本研究所提出之研究方法不受產業與商品限制，具有因應不同使用者之背景與未來應用產業的彈性。
關鍵字: 資料探勘;需求預測;商品分類;基因演算法;供應鏈管理;
Demand Forecasting Using Data Mining Aided Product Classification
Abstract: Product classification is the core of every information activity related to product management. Companies classify their products according to some attributes for different management functions. Within these functions, demand forecasting is the most critical oneof businessmanagement. Forecasts are essential to the business's decision making and planning processes. Better forecasting can contribute to better price structuring and better inventory management. However, it is a challenging problem owing to the volatility of demand which depends on many factors. Therefore, the study aims to design a classification scheme based on the quantitative characteristics of products and make it more suitable for demand forecasting. This study proposes a heuristic algorithm, called Data-Mining Aided Product Classification (DMAPC), to deal with aforementioned issues. DMAPC analyzes the sales records by using time-series analysis and searches the optimal product grouping result using GA-based heuristic algorithm. Accordingly, a new classification scheme is constructed by aforementioned processes. The proposed approach is applied to solve two real-world demand forecasting problems from a well-known cosmetic chain retailer and a prestigious tea retailer in Taiwan. The experimental results demonstrate that the proposed approach is proved to enhance the prediction accuracy effectively by applying DMAPC.
Keywords: Data Mining;Demand Forecasting;Product Classification;Genetic Algorithm;Supply Chain Management;
瀏覽次數: 16808 下載次數: 3782
引用 導入Endnote