<< Ordinal Methods | index | Outranking Methods >>

Introduction to decision aid

Multi-Attribute Utility methods

The decision aid methods need to aggregate the decision maker's preferences somehow. The weighted sum method does it in an intuitive way. Unfortunately, the reality does not always properly fit in such basic mathematical models.

In particular, it happens often that the decision maker cannot determine to what extend she is prepared to sacrifice on one criteria for a given improvement of another : when a valley must be flooded to build a dam, how many more houses can you accept to flood for n more megawatts ? This is called the trade-off, and needs to be taken in account in methods that aim to model the reality closer. This implies quite complex mathematical developments, well beyond the problem we already mentioned, which can be caused by the normalization methods.

The decision aid methods called multi-attribute utility methods (like UTA or MAUT) cover these aspects their own way, but have in common that they assume that the decision maker's preferences are cumulative. They deduct from the decision matrix the utility functions for the criteria, and then solve the equations system that maximizes the results.

At this stage, DECIDE does not include yet functions directly related to these methods.

<< Ordinal Methods | index | Outranking Methods >>


printer friendly view

Google

 

Web

www.christophedavid.org