帳號:guest(13.58.217.242)          離開系統
字體大小: 字級放大   字級縮小   預設字形  

詳目顯示

以作者查詢圖書館館藏以作者查詢臺灣博碩士論文系統以作者查詢全國書目
作者(中文):邱于晅
作者(外文):Chiu, Yu-Hsuan
論文名稱(中文):利用實驗室測試資料來預測產品在保固期內之退貨率
論文名稱(外文):Field Return Rate Prediction within Warranty Period (Based on Laboratory Testing Data)
指導教授(中文):曾勝滄
指導教授(外文):Tseng, Sheng-Tsaing
口試委員(中文):徐南蓉
汪上曉
彭健育
口試委員(外文):Hsu, Nan-Jung
Wong, Shan-Hill
Peng, Chien-Yu
學位類別:碩士
校院名稱:國立清華大學
系所名稱:統計學研究所
學號:105024504
出版年(民國):107
畢業學年度:106
語文別:中文
論文頁數:38
中文關鍵詞:保固期可靠度層狀結構退貨率預測
外文關鍵詞:WarrantyReliabilityHierarchical SturctureReturn Rate Prediction
相關次數:
  • 推薦推薦:0
  • 點閱點閱:565
  • 評分評分:*****
  • 下載下載:17
  • 收藏收藏:0
近年來,由於消費者意識抬頭,購買產品時大多會關注產品的製造商是否有提供保固期 (warranty period) 之服務保障,亦即當產品在保固期內失效時,廠商須提供消費者免費維修或更換新產品的服務。故廠商在保固期內,需要準備充足的備料以因應消費者有瑕疵產品之更換,然而,過多的庫存量,將造成庫存成本及費用增加,故如何精準地評估產品在保固期內之退貨率是廠商的重要決策課題。林逸樵 (2013) 針對 “通過/不通過” 之多產品抗壓力及耐摔性測試之實驗室資料建立“層狀結構”模型,而其現場退貨資料則使用limited failure population (LFP) 模型來配適,並透過此兩部分模型之參數連接來建立連結函數。雖然此連結函數有不錯的預測效果,但是較難解釋且有過度擬合 (over-fitting) 的問題。為克服此困難,本研究首先將實驗室中屬性資料 (attribute data) 之未知參數採用羅吉斯模型 (logistic model) 來表示,其優點是可有效地降低實驗室模型之參數數量。同時,透過層狀結構方法可建構出退貨率的預測模型。除此之外,當實驗室資料中完全無失效之情況,本文亦提出補值後的層狀結構分析方法。由實證資料可發現,本研究提出之方法對於現場退貨率之預測能力及連結函數之可解釋性皆有顯著提升效果。值得一提的是,從分析結果可以發現只需要使用補值後之耐摔性測試資料,便能達到良好的退貨率預測效果,此發現對於建構及蒐集適當的實驗室資料將有實質助益。
In recent years, with the rise of consumer awareness, most of the consumers will concern about the warranty policy provided by the product manufacturers. That is, in the warranty period, the manufacturer shall provide consumers with free repair or replacement of new products. On the other hand, the manufacturers need to prepare sufficient stock for replacing defective products in the warranty period. However, excessive inventory will result in inventory costs substantially. Therefore, how to accurately assess the product’s field return rate within the warranty period is a major task to the manufactures. Recently, Lin (2013) proposed a hierarchical structure for modeling the laboratory "go/no go" data, while the field return data were adapted by using the limited failure population (LFP) model. A linkage function is established through the parameter connection of the two datasets. Although this linkage has a good predictive performance, it does suffer from the difficulty of explaining the physical relationship between laboratory and field data. In addition, it also suffers from an over-fitting problem. In order to overcome this difficulty, this thesis first expresses the unknown parameters in the laboratory attribute data in terms of a logistic model. It has the advantages that the number of parameters of the underlying laboratory model can be effectively reduced. In addition, this thesis also suggests an imputation method for the case of no failure observations in the laboratory testing. From the real case study, we found that even though the new approach only uses drop testing data outperforms the approach of Lin (2013).
摘要
目錄
第一章 緒論 ---------------------- 1
1.1 前言 ---------------------- 1
1.2 研究動機與目的 ----------------- 2
1.3 研究架構 -------------------- 3
第二章 文獻探討與問題描述 --------------- 4
2.1 文獻回顧 -------------------- 4
2.2 實驗室失效模型及層狀結構 ------------ 6
2.3 現場失效模型 ------------------ 8
2.4 連結方法及預測績效 --------------- 9
2.5 問題描述 -------------------- 11
第三章 模型建構及退貨率之預測分析 ----------- 13
3.1 實驗室測試資料模型 --------------- 13
3.1.1. 步驟一:最大概似函數估計法 ----------- 14
3.1.2. 步驟二:連續型層狀結構估計法 ---------- 15
3.1.3. 步驟三 --------------------- 17
3.1.4. 預測現場失效模型之方法 ------------- 18
3.2 實例說明 -------------------- 19
3.2.1. Lab A 抗壓力測試 ---------------- 19
3.2.2. Lab B 耐摔性測試 ---------------- 21
3.2.3. 建立連結函數 ------------------ 22
3.2.4. 本研究與林逸樵 (2013) 之預測績效比較 ------ 25
3.3 實驗室測試無失效資料之補值及其分析 ------- 28
3.3.1. 實驗室資料之補值 ---------------- 28
3.3.2. 實驗室失效模型之配適結果 ------------ 30
3.3.3. 本研究與林逸樵 (2013) 之預測績效比較 ------ 32
第四章 結論與後續研究 ----------------- 36
參考文獻 ------------------------- 38

[1] 林逸樵 (2013). “品質測試資料應用於保固產品之退貨率預測”, 國立清華大學統計學研究所碩士論文。
[2] Agresti, A., and Hitchcock, D. B. (2005). “Bayesian inference for categorical data analysis,” Statistical Methods and Applications, 14, 297-330.
[3] Casella, G., and Berger, R.L. (1990). Statistical Inference, Wadsworth and Brooks/Cole, Pacific Grove, CA.
[4] Hong, Y., and Meeker, W. Q. (2010). “Field-failure and warranty prediction based on auxiliary use-rate information,” Technometrics, 52, 148-159.
[5] Meeker, W. Q. (1987). “Limited failure population life tests: application to integrated circuit reliability,” Technometrics, 29, 51-65.
[6] Meeker, W. Q., Escobar, L. A., and Hong, Y. (2009). “Using accelerated life tests results to predict product field reliability,” Technometrics, 51, 146-161.
[7] Suzuki, K. (1985). “Estimation of lifetime parameters from incomplete field data. Technometrics,” 27, 263-271.
[8] Tseng, S. T., Hsu, N. J., and Lin Y. C. (2016). “Joint modeling of laboratory and field data with application to warranty prediction for highly reliable products." IIE Transactions, 48, 710-719.
 
 
 
 
第一頁 上一頁 下一頁 最後一頁 top
* *