0
$\begingroup$

I would like to know if is there any theory about optimization when the objective function is not unique. For example, I would like to maximize the sum of utilities of some agents but also I would like to minimize the sum of the prices that the agents pay.

It is like find a trade-off between two or more things. Is there any optimization theory like this? Do you know any good paper that deals with such trade-off of maximizing something and minimizing another thing?

$\endgroup$
  • $\begingroup$ Then maximize utility minus cost. $\endgroup$ – Rodrigo de Azevedo Aug 3 '16 at 17:47
  • $\begingroup$ The set of all potentially optimal tradeoffs between multiple objectives is called the Pareto frontier. See also multi-objective optimization. $\endgroup$ – user856 Aug 3 '16 at 17:59
1
$\begingroup$

In an optimization problem you need one objective function. For example, in the usual profit maximization problem

$$\max_{x\geq 0}\pi(x)$$

where $\pi$ is our profit/objective function. Letting $R$ be the revenue function and $C$

$$\pi(x)=R(x)-C(x)$$

so maximizing profit is maximizing the sum of revenue and negative cost. More generally, we could incorporate the tradeoff between two objective using a weighted sum of the two objective functions $g$ and $h$:

$$f(x)=\lambda g(x)+(1-\lambda)h(x)$$

where $\lambda\in[0,1]$. In the profit maximization example, $\lambda=1/2$, $g(x)=2R(x)$, and $h(x)=-2C(x)$.

For the example mentioned in the OP, $g$ could be the sum of utilities, and $h$ the negative of the sum of prices paid as a function of the choice variable $x$.

But you may want to think about whether what you want to do makes sense in terms of the theory.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.