以模糊多目標決策方法定義商用筆記型電腦之S型效用函數

Abstract

近年來，筆記型電腦在受到平板電腦和手機的衝擊下，銷量逐年下降，從近年的電腦銷量來看，整體和消費型市場不斷萎縮，相反的，商用筆記型電腦則維持和歷年來一樣的需求總量，對於商用筆記型電腦在商務採購人士的需求，以前都追求穩定、可靠、安全、可管理，但近幾年，商用筆記型電腦有朝向高性能化和輕薄化方向發展的趨勢；商務人士在採買筆記型電腦時，常常無法得到足夠的量化資訊，根據商務採購人士的需求背景不同，和商務採購者的風險態度也影響採買條件的差異，本研究運用多目標決策方法(Multi-Objective Decision-Making, MODM)，結合效用函數和風險態度來協助商務採購人士做出購入商用筆記型電腦的偏好選擇。先從文獻探討中，找出其商用筆記型電腦的選購準則，然後根據商務採購人士的偏好，建立專家問卷，歸納出各準則的效用函數及各準則的重要程度，但是由於資訊的不充足或選購者往往有個人主觀喜好，所以選購的時候，常常無法準確的了解其真正需求的產品，而當產生區間權重時，有些情況會發生權重上界和小於1，或是下界大於1的不合理情形發生，此時採用區間模糊寬鬆縮減 (Interval Fuzzy Leniency Reduction) 的方法來處理區間權重值，運用效用函數和商務採購者的風險態度結合做探討，來套入文中的案例，最後使用目標規劃法 (Goal Programming, GP) 在S型效用函數 (S-Shaped Utility Function) 上進行求解驗證，在各個準則目標交互的運算下，求出最接近目標值的滿意解，獲得效用滿意度最高的商用筆記型電腦作為最終的決策結果。The market of laptop computer is declining by mobile phone and tablet influenced in recent year. Based on the selling quantity, the market of consumer laptop continue to shrink. Instead, the commercial laptop maintains the shipment quantity. The demand of commercial is used to focus on stability, reliability, security and management, but the criteria add functionality and thin& light feature additionally in the recent. But while business buyers selecting commercial laptop computers, it is hard to get the quantized data of commercial laptop computer. Furthermore, based on the different background and preference on a business buyer, the utility and requirement will be different. The methodologies in this study are using Multi-Objective Decision-Making, MODM to help selecting commercial laptop. This study adopt S-Shaped Utility Function with risk attitude to represent business buyer`s preferences. First, review the literature to define the aspects and criteria. Second, create the utility function and the weight of criteria. Because of insufficient information, people cannot precisely select the product. The weight of criteria might affect the selection. Therefore, the interval values represent the preference of business buyer. The sum of weights usually have an unreasonable status that the upper boundary becomes less than 1, or the lower boundary becomes greater than 1. To avoid this problem, interval fuzzy leniency reduction is applied in this study to overcome the problem which mentioned. Final, using the goal programming to derive the S-Shaped utility function with risk attitude, it can help business buyer to select optimal alternative.