Optimization with parameters - MathOverflow most recent 30 from http://mathoverflow.net 2013-05-25T21:35:00Z http://mathoverflow.net/feeds/question/118402 http://www.creativecommons.org/licenses/by-nc/2.5/rdf http://mathoverflow.net/questions/118402/optimization-with-parameters Optimization with parameters MLT 2013-01-08T23:04:03Z 2013-01-08T23:22:34Z <p>I'm trying to formulate a game so that at Nash equilibrium I achieve supply equales demand. Then I ran into this problem.</p> <p>For all $i,$ $v_{i}\left(x_{i}\right)$ is concave in $x_{i}$. The value function of user $i.$ For all $k$ $c_{k}\left(y_{k}\right)$ is convex in $y_{k}$ and non negative. The cost function of producer $k.$ So that $\sum_{i}v_{i}\left(x_{i}\right)-\sum_{k}c_{k}\left(y_{k}\right)$ is concave.</p> <p>Let $x=\left(x_{1},\,\cdots x_{N}\right)$ and $y=\left(y_{1},\cdots y_{M}\right).$</p> <p>Let price per unit be</p> <p>$$p=a_{0}+a_{1}\sum_{i}x_{i}+a_{2}\sum_{k}y_{k}.$$</p> <p>And the objective, which I have verified is an exact potential function of the game, is</p> <p>$$\Phi\left(x,\, y\right)=\sum_{i}v_{i}\left(x_{i}\right)-\sum_{k}c_{k}\left(y_{k}\right)-p\sum_{i}x_{i}+p\sum_{k}y_{k}.$$</p> <p>Show that under some assumptions $\exists\, a_{0},\, a_{1},\, a_{2}$ such that if,</p> <p>$$(x^{'}, y^{'})=\underset{y_{k}\in Y_{k},\, x_{i}\in X_{i}}{\text{arg max }}\Phi\left(x, y\right)$$</p> <p>then $\sum_{i\in I}x_{i}^{'}=\sum_{k\in K}y_{k}^{'}$. </p> <p>I think this statement is equivalent to showing $\exists\, a_{0},\, a_{1},\, a_{2}$ such that if,</p> <p>$$\left(x^{"},y^{"}\right)=\underset{y_{k}\in Y_{k},\, x_{i}\in X_{i}}{\text{arg max }}\Phi\left(x,\, y\right).$$</p> <p>and if $$\left(x^{'}, y^{'}\right)=\underset{y_{k}\in Y_{k},\, x_{i}\in X_{i},\,\sum_{i\in I}x_{i}=\sum_{k\in K}y_{k}}{\text{arg max }}\Phi\left(x,\, y\right)$$</p> <p>then</p> <p>$\left(x^{'}, y^{'}\right)=\left(x^{"}, y^{"} \right)$</p> <p>For the case when $a_{1}=a_{2}=0$ I have proved that $a_0$ is the Lagrange multiplier but I Can't solve for this general case.</p>