UtilityWerte

UtilityWerte, often translated as utility values, are numerical representations of the level of satisfaction, preference, or value that an agent assigns to possible outcomes. In decision theory and economics, they are used to model preferences in a way that allows comparison and mathematical analysis of choices.

A utility function u maps outcomes or prospects to a real number. When a decision is made under uncertainty, the framework of expected utility suggests choosing the option that maximizes the expected utility E[u(X)]. Utility can be cardinal (the numbers carry meaningful differences and ratios) or merely ordinal (only the order matters).

Utility values can be obtained through direct elicitation methods such as standard gamble, time trade-off, or rating scales, and, in some contexts, indirectly through revealed preferences or observed choices. In health economics, for instance, utility values on a 0–1 scale underpin quality-adjusted life years (QALYs). In artificial intelligence, similar concepts appear as reward or value functions used by agents to guide learning, often via reinforcement learning.

Limitations: utility is a representation of preferences and depends on the agent; the numerical scale is inherently idiosyncratic and may not be comparable across individuals. Risk attitudes are captured by the curvature of the utility function, and contextual or cultural factors can influence elicited values. Related concepts include utility theory, expected utility, and multi-attribute utility theory, as well as applications in health economics and reinforcement learning.