NutzensWeights

NutzensWeights are a framework used in decision analysis to express the relative importance of multiple attributes in a single numerical score. They consist of a vector of weights assigned to each attribute, such that the overall utility of an option is the weighted sum of its attribute values. In normalized formulations, the weights sum to one and are nonnegative.

Calculation and elicitation: Weights can be obtained from observed choices via inverse optimization, from stated preferences through structured surveys, or learned from data using machine learning and optimization techniques. The resulting weights reflect the decision maker's marginal utility for each attribute, and may be constrained to preserve monotonicity or scaled to accommodate heterogeneous attribute ranges.

Applications: NutzensWeights are used in multi-criteria decision analysis, portfolio selection, recommender systems, and policy evaluation to produce a single comparable score for diverse options. They support transparent tradeoffs by making explicit the relative importance of each attribute.

Limitations and variants: The additive, linear form assumes attribute independence and comparable scales, which may not hold in all cases. Sensitivity to scaling and preference inconsistencies can affect outcomes. Extensions include nonlinear utility functions, interaction terms, and hierarchical weighting schemes, along with probabilistic or Bayesian approaches to capture uncertainty.

Origin of the term: Nutzens is derived from the German word for utility; NutzensWeights is a coined term used in theoretical discussions to describe utility-based weighting schemes across disciplines.