January 2020
Lexicographic Approach A lexicographic approach allows decisionmakers to introduce decision rules in which they select more objects impacting on their most-preferred criteria. According to Saban and Sethuraman (2014), when two objects have the same impact on the most-preferred criteria, decision makers prefer the one with the highest impact on the second most-preferred criteria, and so forth. This lexicographic representation models the problems where decision makers strictly prefer one criterion over another or they are managing noncompensatory aggregation (Pulido, Mandow, & de la Cruz, 2014; Yaman, Walsh, Littman, & Desjardins, 2011). Finally, decisionmakers canmodel their strong preferences over the criteria selectedmainly because, after further analysis of the problem, they are not indiferent or only weakly sure about their preferences on the criteria taken into consideration. In other words, they will always prefer one criterion to another without considering criterion weights explicitly. Risk Metrics and Compliance Risk metrics are statistical indicators or measurements that allow decision makers to analyze the dispersion (volatility) of certain events or outcomes. Hence, a random variable can be evaluated using statistical moments (e.g., mean, variance, skewness, kurtosis), or risk measurements can be used to analyze extreme values, such as Value at Risk (VaR) and Conditional VaR (Bodie, Kane, &Marcus, 2009; Fabozzi, 2010; Matos, 2007; Mun, 2015). In decision problems, risk metrics play an important role in analyzing the volatility or stability of a set of options or a portfolio of alternatives, for example, in fnancial risk management (Chong, 2004), portfolio risk management (Bodie et al., 2009), and enterprise riskmanagement (Scarlat, Chirita, & Bradea, 2012), as well as a variety of other areas (Fabozzi, 2010; Szolgayová, Fuss, Khabarov, & Obersteiner, 2011). In order to determine how risky an object is and its relationship with other objects, a compliance approach is followed, that is, the defnition of a set of rules to guide decision makers (Hopkins, 2011). Various analysts have proposed several approaches for assessing compliance. For example, Barrett and Donald (2003) propose a stochastic dominance analysis to compare probability distributions before establishing a hierarchy; Boucher, Danielsson, Kouontchou, and Maillet (2014) rely on risk metrics and forecasting to adjust models by historical performance; and Zanoli, Gambelli, Solfanelli, and Padel (2014) analyze impacts of risk factors on noncompliance in UK farming.
81
Defense ARJ, January 2020, Vol. 27No. 1 : 60-107
Made with FlippingBook Learn more on our blog