Fuzzy multiobjective decision making software


Real world decisions are characterized by uncertainty. Often, our preferences for every possible action are vague and not fully driven by rationality.

This comes true, in particular, when we have to deal with serious consequences. Probability theory, we are used to adopting in decision making processes, requires robust information to work properly.

When preliminary information is not strong enough to use probability theory, fuzzy algorythms come useful.

The example we present here (below) is a particular implementation of the min-max algorithm, from which it differs in considering the client preference factors.


The full software allows setting of the number of alternatives and their names plus the number of objectives and their names.


IMPORTANT: these alternatives may be of any kind (they are only defined by quality factor values you choose). More, quality factor types in this example are fixed (cost, performance, flexibility) but users can define their own ones in the full version.


FUZZY OPTIMIZATION WITH PREFERENCE FACTORS

You have to choose between three alternatives: A, B and C (they can be of any kind).

Three objectives impact the decision: cost, performance, flexibility to future changes of the requirements;

so, you must define quality factors for each objective (in the scale from 0.1 (bad) to 0.9 (optimum) ).

Moreover, your client has defined preferences for each objective (in the scale from 0.1 (small preference) to 0.9 (full preference) ).

Having these preconditions, the algorithm must calculate the best solution among the three alternatives.

Each parameter has a certain degree of validity, so this is a typical fuzzy problem.

The three alternatives may be of any kind: you define them by quality factor values only.


A cost
A performance
A flexibility
B cost
B performance
B flexibility
C cost
C performance
C flexibility
client preference: cost
client preference: performance
client preference: flexibility






RESULT




result coeff.