評估權重設定方式之選用分析

The Selection and Analysis of Weighting Methods for Technical Evaluation

王禎麟(Zhen-Lin Wang)

指導教授 ： 傅增棣

國立臺灣大學/工學院/機械工程學研究所/碩士(2009年)

https://doi.org/10.6342/NTU.2009.00287

摘要

評估條件之權重一般是由評估者所賦予，且是影響評估結果符合評估目的之重要因素，也能協助設計工程師預估消費者對於產品屬性之期望，因此本研究嘗試尋找權重估計誤差較小的方法，以提供設計決策時之所需。首先會建立一套評估權重賦予方式的比較基準，以評比各個多條件決策方法，再藉由比較結果，說明各個方法的優缺點與適用時機。在其中，誤差較小的方法為最佳化權重模型法。而權重亦具有跨年變化之特性，因此可將其他可依據歷史數據作預測之時間序列方法納入，顯示自迴歸移動平均模型（即ARMA模型）除了具有估計年份越近，誤差會越小之特性外，在資料補足後，其所估計之誤差亦為最小。因此，可以使用ARMA模型作為評估條件歷年權重演變趨勢之預估。

關鍵字

權重 ； 多條件決策 ； 時間序列方法

並列摘要

Weightings of the evaluation criteria are normally determined by evaluators, which are also important factors influencing results of evaluation. Engineers need to know preference of criteria based on consumer’s viewpoints through weightings. The objective of this research is to find a method to more accurately estimating weightings for evaluation. A standard set of data has been devised in this research for comparing various multiple criteria decision making methods. The comparison shows benefits and drawbacks of each method along with conditions of applications. It concludes that the optimum weight model method is more accurate than others. Weightings also have the properties of changing values over time. Therefore, the time series analysis methods for historical data should be introduced. The autoregressive moving average (ARMA) model was found possessing the properties of better accuracy for near-period weightings, and was more accurate with adequate data. The ARMA model is therefore a better method for estimating the trend weightings in history.

並列關鍵字

weighting ； multiple criteria decision making ； time series analysis method

參考文獻

[14]Belton, V. and Gear, A. E., On a Shortcoming of Saaty’s Method of Analytic Hierarchies, Omega, Vol. 11, No. 3, pp. 227-230, 1983.

[15]Box, G. E. and Jenkins, G. W., Time series Analysis: Forecasting and Control, Holden-Day, San Francisco, 1976.

[16]Brown, R. G., Smoothing, Forecasting and Prediction of Discrete Time Series, Prentice-Hall, New Jersey, 1963.

[17]Brown, R. G., Statistical Forecasting for Inventory Control, McGraw-Hill, New York, 1959.

[18]Chu, A. T. W., Kalaba, R. E. and Spingarn, K., A Comparison of Two Methods for Determining the Weights of Belonging to Fuzzy Sets, Journal of Optimization Theory and Application, Vol. 27, No. 4, pp. 531-538, 1979.